Summary In this episode of the Data Engineering Podcast Ariel Pohoryles, head of product marketing for Boomi's data management offerings, talks about a recent survey of 300 data leaders on how organizations are investing in data to scale AI. He shares a paradox uncovered in the research: while 77% of leaders trust the data feeding their AI systems, only 50% trust their organization's data overall. Ariel explains why truly productionizing AI demands broader, continuously refreshed data with stronger automation and governance, and highlights the challenges posed by unstructured data and vector stores. The conversation covers the need to shift from manual reviews to automated pipelines, the resurgence of metadata and master data management, and the importance of guardrails, traceability, and agent governance. Ariel also predicts a growing convergence between data teams and application integration teams and advises leaders to focus on high-value use cases, aggressive pipeline automation, and cataloging and governing the coming sprawl of AI agents, all while using AI to accelerate data engineering itself.

Announcements Hello and welcome to the Data Engineering Podcast, the show about modern data managementData teams everywhere face the same problem: they're forcing ML models, streaming data, and real-time processing through orchestration tools built for simple ETL. The result? Inflexible infrastructure that can't adapt to different workloads. That's why Cash App and Cisco rely on Prefect. Cash App's fraud detection team got what they needed - flexible compute options, isolated environments for custom packages, and seamless data exchange between workflows. Each model runs on the right infrastructure, whether that's high-memory machines or distributed compute. Orchestration is the foundation that determines whether your data team ships or struggles. ETL, ML model training, AI Engineering, Streaming - Prefect runs it all from ingestion to activation in one platform. Whoop and 1Password also trust Prefect for their data operations. If these industry leaders use Prefect for critical workflows, see what it can do for you at dataengineeringpodcast.com/prefect.Data migrations are brutal. They drag on for months—sometimes years—burning through resources and crushing team morale. Datafold's AI-powered Migration Agent changes all that. Their unique combination of AI code translation and automated data validation has helped companies complete migrations up to 10 times faster than manual approaches. And they're so confident in their solution, they'll actually guarantee your timeline in writing. Ready to turn your year-long migration into weeks? Visit dataengineeringpodcast.com/datafold today for the details.Composable data infrastructure is great, until you spend all of your time gluing it together. Bruin is an open source framework, driven from the command line, that makes integration a breeze. Write Python and SQL to handle the business logic, and let Bruin handle the heavy lifting of data movement, lineage tracking, data quality monitoring, and governance enforcement. Bruin allows you to build end-to-end data workflows using AI, has connectors for hundreds of platforms, and helps data teams deliver faster. Teams that use Bruin need less engineering effort to process data and benefit from a fully integrated data platform. Go to dataengineeringpodcast.com/bruin today to get started. And for dbt Cloud customers, they'll give you $1,000 credit to migrate to Bruin Cloud.Your host is Tobias Macey and today I'm interviewing Ariel Pohoryles about data management investments that organizations are making to enable them to scale AI implementationsInterview IntroductionHow did you get involved in the area of data management?Can you start by describing the motivation and scope of your recent survey on data management investments for AI across your respondents?What are the key takeaways that were most significant to you?The survey reveals a fascinating paradox: 77% of leaders trust the data used by their AI systems, yet only half trust their organization's overall data quality. For our data engineering audience, what does this suggest about how companies are currently sourcing data for AI? Does it imply they are using narrow, manually-curated "golden datasets," and what are the technical challenges and risks of that approach as they try to scale?The report highlights a heavy reliance on manual data quality processes, with one expert noting companies feel it's "not reliable to fully automate validation" for external or customer data. At the same time, maturity in "Automated tools for data integration and cleansing" is low, at only 42%. What specific technical hurdles or organizational inertia are preventing teams from adopting more automation in their data quality and integration pipelines?There was a significant point made that with generative AI, "biases can scale much faster," making automated governance essential. From a data engineering perspective, how does the data management strategy need to evolve to support generative AI versus traditional ML models? What new types of data quality checks, lineage tracking, or monitoring for feedback loops are required when the model itself is generating new content based on its own outputs?The report champions a "centralized data management platform" as the "connective tissue" for reliable AI. How do you see the scale and data maturity impacting the realities of that effort?How do architectural patterns in the shape of cloud warehouses, lakehouses, data mesh, data products, etc. factor into that need for centralized/unified platforms?A surprising finding was that a third of respondents have not fully grasped the risk of significant inaccuracies in their AI models if they fail to prioritize data management. In your experience, what are the biggest blind spots for data and analytics leaders?Looking at the maturity charts, companies rate themselves highly on "Developing a data management strategy" (65%) but lag significantly in areas like "Automated tools for data integration and cleansing" (42%) and "Conducting bias-detection audits" (24%). If you were advising a data engineering team lead based on these findings, what would you tell them to prioritize in the next 6-12 months to bridge the gap between strategy and a truly scalable, trustworthy data foundation for AI?The report states that 83% of companies expect to integrate more data sources for their AI in the next year. For a data engineer on the ground, what is the most important capability they need to build into their platform to handle this influx?What are the most interesting, innovative, or unexpected ways that you have seen teams addressing the new and accelerated data needs for AI applications?What are some of the noteworthy trends or predictions that you have for the near-term future of the impact that AI is having or will have on data teams and systems?Contact Info LinkedInParting Question From your perspective, what is the biggest gap in the tooling or technology for data management today?Closing Announcements Thank you for listening! Don't forget to check out our other shows. Podcast.init covers the Python language, its community, and the innovative ways it is being used. The AI Engineering Podcast is your guide to the fast-moving world of building AI systems.Visit the site to subscribe to the show, sign up for the mailing list, and read the show notes.If you've learned something or tried out a project from the show then tell us about it! Email [email protected] with your story.Links BoomiData ManagementIntegration & Automation DemoAgentstudioData Connector Agent WebinarSurvey ResultsData GovernanceShadow ITPodcast EpisodeThe intro and outro music is from The Hug by The Freak Fandango Orchestra / CC BY-SA

Summary The data ecosystem has been growing rapidly, with new communities joining and bringing their preferred programming languages to the mix. This has led to inefficiencies in how data is stored, accessed, and shared across process and system boundaries. The Arrow project is designed to eliminate wasted effort in translating between languages, and Voltron Data was created to help grow and support its technology and community. In this episode Wes McKinney shares the ways that Arrow and its related projects are improving the efficiency of data systems and driving their next stage of evolution.

Announcements

Hello and welcome to the Data Engineering Podcast, the show about modern data management When you’re ready to build your next pipeline, or want to test out the projects you hear about on the show, you’ll need somewhere to deploy it, so check out our friends at Linode. With their new managed database service you can launch a production ready MySQL, Postgres, or MongoDB cluster in minutes, with automated backups, 40 Gbps connections from your application hosts, and high throughput SSDs. Go to dataengineeringpodcast.com/linode today and get a $100 credit to launch a database, create a Kubernetes cluster, or take advantage of all of their other services. And don’t forget to thank them for their continued support of this show! Atlan is the metadata hub for your data ecosystem. Instead of locking your metadata into a new silo, unleash its transformative potential with Atlan’s active metadata capabilities. Push information about data freshness and quality to your business intelligence, automatically scale up and down your warehouse based on usage patterns, and let the bots answer those questions in Slack so that the humans can focus on delivering real value. Go to dataengineeringpodcast.com/atlan today to learn more about how Atlan’s active metadata platform is helping pioneering data teams like Postman, Plaid, WeWork & Unilever achieve extraordinary things with metadata and escape the chaos. Struggling with broken pipelines? Stale dashboards? Missing data? If this resonates with you, you’re not alone. Data engineers struggling with unreliable data need look no further than Monte Carlo, the leading end-to-end Data Observability Platform! Trusted by the data teams at Fox, JetBlue, and PagerDuty, Monte Carlo solves the costly problem of broken data pipelines. Monte Carlo monitors and alerts for data issues across your data warehouses, data lakes, dbt models, Airflow jobs, and business intelligence tools, reducing time to detection and resolution from weeks to just minutes. Monte Carlo also gives you a holistic picture of data health with automatic, end-to-end lineage from ingestion to the BI layer directly out of the box. Start trusting your data with Monte Carlo today! Visit dataengineeringpodcast.com/montecarlo to learn more. Data engineers don’t enjoy writing, maintaining, and modifying ETL pipelines all day, every day. Especially once they realize 90% of all major data sources like Google Analytics, Salesforce, Adwords, Facebook, Spreadsheets, etc., are already available as plug-and-play connectors with reliable, intuitive SaaS solutions. Hevo Data is a highly reliable and intuitive data pipeline platform used by data engineers from 40+ countries to set up and run low-latency ELT pipelines with zero maintenance. Boasting more than 150 out-of-the-box connectors that can be set up in minutes, Hevo also allows you to monitor and control your pipelines. You get: real-time data flow visibility, fail-safe mechanisms, and alerts if anything breaks; preload transformations and auto-schema mapping precisely control how data lands in your destination; models and workflows to transform data for analytics; and reverse-ETL capability to move the transformed data back to your business software to inspire timely action. All of this, plus its transparent pricing and 24*7 live support, makes it consistently voted by users as the Leader in the Data Pipeline category on review platforms like G2. Go to dataengineeringpodcast.com/hevodata and sign up for a free 14-day trial that also comes with 24×7 support. Your host is Tobias Macey and today I’m interviewing Wes McKinney about his work at Voltron Data and on the Arrow ecosystem

Interview

Introduction How did you get involved in the area of data management? Can you describe what you are building at Voltron Data and the story behind it? What is the vision for the broader data ecosystem that you are trying to realize through your investment in Arrow and related projects?

How does your work at Voltron Data contribute to the realization of that vision?

What is the impact on engineer productivity and compute efficiency that gets introduced by the impedance mismatches between language and framework representations of data? The scope and capabilities of the Arrow project have grown substantially since it was first introduced. Can you give an overview of the current features and extensions to the project? What are some of the ways that ArrowVe and its related projects can be integrated with or replace the different elements of a data platform? Can you describe how Arrow is implemented?

What are the most complex/challenging aspects of the engineering needed to support interoperable data interchange between language runtimes?

How are you balancing the desire to move quickly and improve the Arrow protocol and implementations, with the need to wait for other players in the ecosystem (e.g. database engines, compute frameworks, etc.) to add support? With the growing application of data formats such as graphs and vectors, what do you see as the role of Arrow and its ideas in those use cases? For workflows that rely on integrating structured and unstructured data, what are the options for interaction with non-tabular data? (e.g. images, documents, etc.) With your support-focused business model, how are you approaching marketing and customer education to make it viable and scalable? What are the most interesting, innovative, or unexpected ways that you have seen Arrow used? What are the most interesting, unexpected, or challenging lessons that you have learned while working on Arrow and its ecosystem? When is Arrow the wrong choice? What do you have planned for the future of Arrow?

Contact Info

Website wesm on GitHub @wesmckinn on Twitter

Parting Question

From your perspective, what is the biggest gap in the tooling or technology for data management today?

Closing Announcements

Thank you for listening! Don’t forget to check out our other shows. Podcast.init covers the Python language, its community, and the innovative ways it is being used. The Machine Learning Podcast helps you go from idea to production with machine learning. Visit the site to subscribe to the show, sign up for the mailing list, and read the show notes. If you’ve learned something or tried out a project from the show then tell us about it! Email [email protected]) with your story. To help other people find the show please leave a review on Apple Podcasts and tell your friends and co-workers

Links

Voltron Data Pandas

Podcast Episode

Apache Arrow Partial Differential Equation FPGA == Field-Programmable Gate Array GPU == Graphics Processing Unit Ursa Labs Voltron (cartoon) Feature Engineering PySpark Substrait Arrow Flight Acero Arrow Datafusion Velox Ibis SIMD == Single Instruction, Multiple Data Lance DuckDB

Podcast Episode

Data Threads Conference Nano-Arrow Arrow ADBC Protocol Apache Iceberg

Podcast Episode

The intro and outro music is from The Hug by The Freak Fandango Orchestra / CC BY-SA

Sponsored By: Atlan:

Have you ever woken up to a crisis because a number on a dashboard is broken and no one knows why? Or sent out frustrating slack messages trying to find the right data set? Or tried to understand what a column name means?

Our friends at Atlan started out as a data team themselves and faced all this collaboration chaos themselves, and started building Atlan as an internal tool for themselves. Atlan is a collaborative workspace for data-driven teams, like Github for engineering or Figma for design teams. By acting as a virtual hub for data assets ranging from tables and dashboards to SQL snippets & code, Atlan enables teams to create a single source of truth for all their data assets, and collaborate across the modern data stack through deep integrations with tools like Snowflake, Slack, Looker and more.

Go to dataengineeringpodcast.com/atlan and sign up for a free trial. If you’re a data engineering podcast listener, you get credits worth $3000 on an annual subscription.a href="https://dataengineeringpodcast.com/montecarlo"…

Summary Reverse ETL is a product category that evolved from the landscape of customer data platforms with a number of companies offering their own implementation of it. While struggling with the work of automating data integration workflows with marketing, sales, and support tools Brian Leonard accidentally discovered this need himself and turned it into the open source framework Grouparoo. In this episode he explains why he decided to turn these efforts into an open core business, how the platform is implemented, and the benefits of having an open source contender in the landscape of operational analytics products.

Announcements

Hello and welcome to the Data Engineering Podcast, the show about modern data management When you’re ready to build your next pipeline, or want to test out the projects you hear about on the show, you’ll need somewhere to deploy it, so check out our friends at Linode. With their managed Kubernetes platform it’s now even easier to deploy and scale your workflows, or try out the latest Helm charts from tools like Pulsar and Pachyderm. With simple pricing, fast networking, object storage, and worldwide data centers, you’ve got everything you need to run a bulletproof data platform. Go to dataengineeringpodcast.com/linode today and get a $100 credit to try out a Kubernetes cluster of your own. And don’t forget to thank them for their continued support of this show! StreamSets DataOps Platform is the world’s first single platform for building smart data pipelines across hybrid and multi-cloud architectures. Build, run, monitor and manage data pipelines confidently with an end-to-end data integration platform that’s built for constant change. Amp up your productivity with an easy-to-navigate interface and 100s of pre-built connectors. And, get pipelines and new hires up and running quickly with powerful, reusable components that work across batch and streaming. Once you’re up and running, your smart data pipelines are resilient to data drift. Those ongoing and unexpected changes in schema, semantics, and infrastructure. Finally, one single pane of glass for operating and monitoring all your data pipelines. The full transparency and control you desire for your data operations. Get started building pipelines in minutes for free at dataengineeringpodcast.com/streamsets. The first 10 listeners of the podcast that subscribe to StreamSets’ Professional Tier, receive 2 months free after their first month. Atlan is a collaborative workspace for data-driven teams, like Github for engineering or Figma for design teams. By acting as a virtual hub for data assets ranging from tables and dashboards to SQL snippets & code, Atlan enables teams to create a single source of truth for all their data assets, and collaborate across the modern data stack through deep integrations with tools like Snowflake, Slack, Looker and more. Go to dataengineeringpodcast.com/atlan today and sign up for a free trial. If you’re a data engineering podcast listener, you get credits worth $3000 on an annual subscription Your host is Tobias Macey and today I’m interviewing Brian Leonard about Grouparoo, an open source framework for managing your reverse ETL pipelines

Interview

Introduction How did you get involved in the area of data management? Can you describe what Grouparoo is and the story behind it? What are the core requirements for building a reverse ETL system?

What are the additional capabilities that users of the system ask for as they get more advanced in their usage?

Who is your target user for Grouparoo and how does that influence your priorities on feature development and UX design? What are the benefits of building an open source core for a reverse ETL platform as compared to the other commercial options? Can you describe the architecture and implementation of the Grouparoo project?

What are the additional systems that you have built to support the hosted offering? How have the design and goals of the

Summary The technological and social ecosystem of data engineering and data management has been reaching a stage of maturity recently. As part of this stage in our collective journey the focus has been shifting toward operation and automation of the infrastructure and workflows that power our analytical workloads. It is an encouraging sign for the industry, but it is still a complex and challenging undertaking. In order to make this world of DataOps more accessible and manageable the team at Nexla has built a platform that decouples the logical unit of data from the underlying mechanisms so that you can focus on the problems that really matter to your business. In this episode Saket Saurabh (CEO) and Avinash Shahdadpuri (CTO) share the story behind the Nexla platform, discuss the technical underpinnings, and describe how their concept of a Nexset simplifies the work of building data products for sharing within and between organizations.

Announcements

Hello and welcome to the Data Engineering Podcast, the show about modern data management When you’re ready to build your next pipeline, or want to test out the projects you hear about on the show, you’ll need somewhere to deploy it, so check out our friends at Linode. With their managed Kubernetes platform it’s now even easier to deploy and scale your workflows, or try out the latest Helm charts from tools like Pulsar and Pachyderm. With simple pricing, fast networking, object storage, and worldwide data centers, you’ve got everything you need to run a bulletproof data platform. Go to dataengineeringpodcast.com/linode today and get a $100 credit to try out a Kubernetes cluster of your own. And don’t forget to thank them for their continued support of this show! Schema changes, missing data, and volume anomalies caused by your data sources can happen without any advanced notice if you lack visibility into your data-in-motion. That leaves DataOps reactive to data quality issues and can make your consumers lose confidence in your data. By connecting to your pipeline orchestrator like Apache Airflow and centralizing your end-to-end metadata, Databand.ai lets you identify data quality issues and their root causes from a single dashboard. With Databand.ai, you’ll know whether the data moving from your sources to your warehouse will be available, accurate, and usable when it arrives. Go to dataengineeringpodcast.com/databand to sign up for a free 30-day trial of Databand.ai and take control of your data quality today. We’ve all been asked to help with an ad-hoc request for data by the sales and marketing team. Then it becomes a critical report that they need updated every week or every day. Then what do you do? Send a CSV via email? Write some Python scripts to automate it? But what about incremental sync, API quotas, error handling, and all of the other details that eat up your time? Today, there is a better way. With Census, just write SQL or plug in your dbt models and start syncing your cloud warehouse to SaaS applications like Salesforce, Marketo, Hubspot, and many more. Go to dataengineeringpodcast.com/census today to get a free 14-day trial. Your host is Tobias Macey and today I’m interviewing Saket Saurabh and Avinash Shahdadpuri about Nexla, a platform for powering data operations and sharing within and across businesses

Interview

Introduction How did you get involved in the area of data management? Can you describe what Nexla is and the story behind it? What are the major problems that Nexla is aiming to solve?

What are the components of a data platform that Nexla might replace?

What are the use cases and benefits of being able to publish data sets for use outside and across organizations? What are the different elements involved in implementing DataOps? How is the Nexla platform implemented?

What have been the most comple engineering challenges? How has the architecture changed or evolved since you first began working on it? What are some of the assumpt

Summary A major concern that comes up when selecting a vendor or technology for storing and managing your data is vendor lock-in. What happens if the vendor fails? What if the technology can’t do what I need it to? Compilerworks set out to reduce the pain and complexity of migrating between platforms, and in the process added an advanced lineage tracking capability. In this episode Shevek, CTO of Compilerworks, takes us on an interesting journey through the many technical and social complexities that are involved in evolving your data platform and the system that they have built to make it a manageable task.

Announcements

Hello and welcome to the Data Engineering Podcast, the show about modern data management When you’re ready to build your next pipeline, or want to test out the projects you hear about on the show, you’ll need somewhere to deploy it, so check out our friends at Linode. With their managed Kubernetes platform it’s now even easier to deploy and scale your workflows, or try out the latest Helm charts from tools like Pulsar and Pachyderm. With simple pricing, fast networking, object storage, and worldwide data centers, you’ve got everything you need to run a bulletproof data platform. Go to dataengineeringpodcast.com/linode today and get a $100 credit to try out a Kubernetes cluster of your own. And don’t forget to thank them for their continued support of this show! Schema changes, missing data, and volume anomalies caused by your data sources can happen without any advanced notice if you lack visibility into your data-in-motion. That leaves DataOps reactive to data quality issues and can make your consumers lose confidence in your data. By connecting to your pipeline orchestrator like Apache Airflow and centralizing your end-to-end metadata, Databand.ai lets you identify data quality issues and their root causes from a single dashboard. With Databand.ai, you’ll know whether the data moving from your sources to your warehouse will be available, accurate, and usable when it arrives. Go to dataengineeringpodcast.com/databand to sign up for a free 30-day trial of Databand.ai and take control of your data quality today. We’ve all been asked to help with an ad-hoc request for data by the sales and marketing team. Then it becomes a critical report that they need updated every week or every day. Then what do you do? Send a CSV via email? Write some Python scripts to automate it? But what about incremental sync, API quotas, error handling, and all of the other details that eat up your time? Today, there is a better way. With Census, just write SQL or plug in your dbt models and start syncing your cloud warehouse to SaaS applications like Salesforce, Marketo, Hubspot, and many more. Go to dataengineeringpodcast.com/census today to get a free 14-day trial. Your host is Tobias Macey and today I’m interviewing Shevek about Compilerworks and his work on writing compilers to automate data lineage tracking from your SQL code

Interview

Introduction How did you get involved in the area of data management? Can you describe what Compilerworks is and the story behind it? What is a compiler?

How are you applying compilers to the challenges of data processing systems?

What are some use cases that Compilerworks is uniquely well suited to? There are a number of other methods and systems available for tracking and/or computing data lineage. What are the benefits of the approach that you are taking with Compilerworks? Can you describe the design and implementation of the Compilerworks platform?

How has the system changed or evolved since you first began working on it?

What programming languages and SQL dialects do you currently support?

Which have been the most challenging to work with? How do you handle verification/validation of the algebraic representation of SQL code given the variability of implementations and the flexibility of the specification?

Can you talk through the process of getting Compilerworks

Summary All of the fancy data platform tools and shiny dashboards that you use are pointless if the consumers of your analysis don’t have trust in the answers. Stemma helps you establish and maintain that trust by giving visibility into who is using what data, annotating the reports with useful context, and understanding who is responsible for keeping it up to date. In this episode Mark Grover explains what he is building at Stemma, how it expands on the success of the Amundsen project, and why trust is the most important asset for data teams.

Announcements

Hello and welcome to the Data Engineering Podcast, the show about modern data management When you’re ready to build your next pipeline, or want to test out the projects you hear about on the show, you’ll need somewhere to deploy it, so check out our friends at Linode. With their managed Kubernetes platform it’s now even easier to deploy and scale your workflows, or try out the latest Helm charts from tools like Pulsar and Pachyderm. With simple pricing, fast networking, object storage, and worldwide data centers, you’ve got everything you need to run a bulletproof data platform. Go to dataengineeringpodcast.com/linode today and get a $100 credit to try out a Kubernetes cluster of your own. And don’t forget to thank them for their continued support of this show! RudderStack’s smart customer data pipeline is warehouse-first. It builds your customer data warehouse and your identity graph on your data warehouse, with support for Snowflake, Google BigQuery, Amazon Redshift, and more. Their SDKs and plugins make event streaming easy, and their integrations with cloud applications like Salesforce and ZenDesk help you go beyond event streaming. With RudderStack you can use all of your customer data to answer more difficult questions and then send those insights to your whole customer data stack. Sign up free at dataengineeringpodcast.com/rudder today. We’ve all been asked to help with an ad-hoc request for data by the sales and marketing team. Then it becomes a critical report that they need updated every week or every day. Then what do you do? Send a CSV via email? Write some Python scripts to automate it? But what about incremental sync, API quotas, error handling, and all of the other details that eat up your time? Today, there is a better way. With Census, just write SQL or plug in your dbt models and start syncing your cloud warehouse to SaaS applications like Salesforce, Marketo, Hubspot, and many more. Go to dataengineeringpodcast.com/census today to get a free 14-day trial. Your host is Tobias Macey and today I’m interviewing Mark Grover about his work at Stemma to bring the Amundsen project to a wider audience and increase trust in their data.

Interview

Introduction Can you describe what Stemma is and the story behind it? Can you give me more context into how and why Stemma fits into the current data engineering world? Among the popular tools of today for data warehousing and other products that stitch data together – what is Stemma’s place? Where does it fit into the workflow? How has the explosion in options for data cataloging and discovery influenced your thinking on the necessary feature set for that class of tools? How do you compare to your competitors With how long we have been using data and building systems to analyze it, why do you think that trust in the results is still such a momentous problem? Tell me more about Stemma and how it compares to Amundsen? Can you tell me more about the impact of Stemma/Amundsen to companies that use it? What are the opportunities for innovating on top of Stemma to help organizations streamline communication between data producers and consumers? Beyond the technological capabilities of a data platform, the bigger question is usually the social/organizational patterns around data. How have the "best practices" around the people side of data changed in the recent past?

What are the points of friction that

Summary Data lake architectures have largely been biased toward batch processing workflows due to the volume of data that they are designed for. With more real-time requirements and the increasing use of streaming data there has been a struggle to merge fast, incremental updates with large, historical analysis. Vinoth Chandar helped to create the Hudi project while at Uber to address this challenge. By adding support for small, incremental inserts into large table structures, and building support for arbitrary update and delete operations the Hudi project brings the best of both worlds together. In this episode Vinoth shares the history of the project, how its architecture allows for building more frequently updated analytical queries, and the work being done to add a more polished experience to the data lake paradigm.

Announcements

Hello and welcome to the Data Engineering Podcast, the show about modern data management You listen to this show to learn about all of the latest tools, patterns, and practices that power data engineering projects across every domain. Now there’s a book that captures the foundational lessons and principles that underly everything that you hear about here. I’m happy to announce I collected wisdom from the community to help you in your journey as a data engineer and worked with O’Reilly to publish it as 97 Things Every Data Engineer Should Know. Go to dataengineeringpodcast.com/97things today to get your copy! When you’re ready to build your next pipeline, or want to test out the projects you hear about on the show, you’ll need somewhere to deploy it, so check out our friends at Linode. With their managed Kubernetes platform it’s now even easier to deploy and scale your workflows, or try out the latest Helm charts from tools like Pulsar and Pachyderm. With simple pricing, fast networking, object storage, and worldwide data centers, you’ve got everything you need to run a bulletproof data platform. Go to dataengineeringpodcast.com/linode today and get a $100 credit to try out a Kubernetes cluster of your own. And don’t forget to thank them for their continued support of this show! RudderStack’s smart customer data pipeline is warehouse-first. It builds your customer data warehouse and your identity graph on your data warehouse, with support for Snowflake, Google BigQuery, Amazon Redshift, and more. Their SDKs and plugins make event streaming easy, and their integrations with cloud applications like Salesforce and ZenDesk help you go beyond event streaming. With RudderStack you can use all of your customer data to answer more difficult questions and then send those insights to your whole customer data stack. Sign up free at dataengineeringpodcast.com/rudder today. We’ve all been asked to help with an ad-hoc request for data by the sales and marketing team. Then it becomes a critical report that they need updated every week or every day. Then what do you do? Send a CSV via email? Write some Python scripts to automate it? But what about incremental sync, API quotas, error handling, and all of the other details that eat up your time? Today, there is a better way. With Census, just write SQL or plug in your dbt models and start syncing your cloud warehouse to SaaS applications like Salesforce, Marketo, Hubspot, and many more. Go to dataengineeringpodcast.com/census today to get a free 14-day trial. Your host is Tobias Macey and today I’m interviewing Vinoth Chandar about Apache Hudi, a data lake management layer for supporting fast and incremental updates to your tables.

Interview

Introduction How did you get involved in the area of data management? Can you describe what Hudi is and the story behind it? What are the use cases that it is focused on supporting? There have been a number of alternative table formats introduced for data lakes recently. How does Hudi compare to projects like Iceberg, Delta Lake, Hive, etc.? Can you describe how Hudi is architected?

How have the goals and design of Hudi changed or evolved since you first began working on it? If you were to start the whole project over today, what would you do differently?

Can you talk through the lifecycle of a data record as it is ingested, compacted, and queried in a Hudi deployment? One of the capabilities that is interesting to explore is support for arbitrary record deletion. Can you talk through why this is a challenging operation in data lake architectures?

How does Hudi make that a tractable problem?

What are the data platform components that are needed to support an installation of Hudi? What is involved in migrating an existing data lake to use Hudi?

How would someone approach supporting heterogeneous table formats in their lake?

As someone who has invested a lot of time in technologies for supporting data lakes, what are your thoughts on the tradeoffs of data lake vs data warehouse and the current trajectory of the ecosystem? What are the most interesting, innovative, or unexpected ways that you have seen Hudi used? What are the most interesting, unexpected, or challenging lessons that you have learned while working on Hudi? When is Hudi the wrong choice? What do you have planned for the future of Hudi?

Contact Info

Linkedin Twitter

Parting Question

From your perspective, what is the biggest gap in the tooling or technology for data management today?

Closing Announcements

Thank you for listening! Don’t forget to check out our other show, Podcast.init to learn about the Python language, its community, and the innovative ways it is being used. Visit the site to subscribe to the show, sign up for the mailing list, and read the show notes. If you’ve learned something or tried out a project from the show then tell us about it! Email [email protected]) with your story. To help other people find the show please leave a review on iTunes and tell your friends and co-workers Join the community in the new Zulip chat workspace at dataengineeringpodcast.com/chat

Links

Hudi Docs Hudi Design & Architecture Incremental Processing CDC == Change Data Capture

Podcast Episodes

Oracle GoldenGate Voldemort Kafka Hadoop Spark HBase Parquet Iceberg Table Format

Data Engineering Episode

Hive ACID Apache Kudu

Podcast Episode

Vertica Delta Lake

Podcast Episode

Optimistic Concurrency Control MVCC == Multi-Version Concurrency Control Presto Flink

Podcast Episode

Trino

Podcast Episode

Gobblin LakeFS

Podcast Episode

Nessie

The intro and outro music is from The Hug by The Freak Fandango Orchestra / CC BY-SA

Support Data Engineering Podcast

Summary Everyone expects data to be transmitted, processed, and updated instantly as more and more products integrate streaming data. The technology to make that possible has been around for a number of years, but the barriers to adoption have still been high due to the level of technical understanding and operational capacity that have been required to run at scale. Datastax has recently introduced a new managed offering for Pulsar workloads in the form of Astra Streaming that lowers those barriers and make stremaing workloads accessible to a wider audience. In this episode Prabhat Jha and Jonathan Ellis share the work that they have been doing to integrate streaming data into their managed Cassandra service. They explain how Pulsar is being used by their customers, the work that they have done to scale the administrative workload for multi-tenant environments, and the challenges of operating such a data intensive service at large scale. This is a fascinating conversation with a lot of useful lessons for anyone who wants to understand the operational aspects of Pulsar and the benefits that it can provide to data workloads.

Announcements

Hello and welcome to the Data Engineering Podcast, the show about modern data management You listen to this show to learn about all of the latest tools, patterns, and practices that power data engineering projects across every domain. Now there’s a book that captures the foundational lessons and principles that underly everything that you hear about here. I’m happy to announce I collected wisdom from the community to help you in your journey as a data engineer and worked with O’Reilly to publish it as 97 Things Every Data Engineer Should Know. Go to dataengineeringpodcast.com/97things today to get your copy! When you’re ready to build your next pipeline, or want to test out the projects you hear about on the show, you’ll need somewhere to deploy it, so check out our friends at Linode. With their managed Kubernetes platform it’s now even easier to deploy and scale your workflows, or try out the latest Helm charts from tools like Pulsar and Pachyderm. With simple pricing, fast networking, object storage, and worldwide data centers, you’ve got everything you need to run a bulletproof data platform. Go to dataengineeringpodcast.com/linode today and get a $100 credit to try out a Kubernetes cluster of your own. And don’t forget to thank them for their continued support of this show! RudderStack’s smart customer data pipeline is warehouse-first. It builds your customer data warehouse and your identity graph on your data warehouse, with support for Snowflake, Google BigQuery, Amazon Redshift, and more. Their SDKs and plugins make event streaming easy, and their integrations with cloud applications like Salesforce and ZenDesk help you go beyond event streaming. With RudderStack you can use all of your customer data to answer more difficult questions and then send those insights to your whole customer data stack. Sign up free at dataengineeringpodcast.com/rudder today. We’ve all been asked to help with an ad-hoc request for data by the sales and marketing team. Then it becomes a critical report that they need updated every week or every day. Then what do you do? Send a CSV via email? Write some Python scripts to automate it? But what about incremental sync, API quotas, error handling, and all of the other details that eat up your time? Today, there is a better way. With Census, just write SQL or plug in your dbt models and start syncing your cloud warehouse to SaaS applications like Salesforce, Marketo, Hubspot, and many more. Go to dataengineeringpodcast.com/census today to get a free 14-day trial. Your host is Tobias Macey and today I’m interviewing Prabhat Jha and Jonathan Ellis about Astra Streaming, a cloud-native streaming platform built on Apache Pulsar

Interview

Introduction

How did you get involved in the area of data management?

Can you describe what the Astra platform is and the story behind it?

How does streaming fit into your overall product vision and the needs of your customers?

What was your selection process/criteria for adopting a streaming engine to complement your existing technology investment?

What are the core use cases that you are aiming to support with Astra Streaming?

Can you describe the architecture and automation of your hosted platform for Pulsar?

What are the integration points that you have built to make it work well with Cassandra?

What are some of the additional tools that you have added to your distribution of Pulsar to simplify operation and use?

What are some of the sharp edges that you have had to sand down as you have scaled up your usage of Pulsar?

What is the process for someone to adopt and integrate with your Astra Streaming service?

How do you handle migrating existing projects, particularly if they are using Kafka currently?

One of the capabilities that you highlight on the product page for Astra Streaming is the ability to execute machine learning workflows on data in flight. What are some of the supporting systems that are necessary to power that workflow?

What are the capabilities that are built into Pulsar that simplify the operational aspects of streaming ML?

What are the ways that you are engaging with and supporting the Pulsar community?

What are the near to medium term elements of the Pulsar roadmap that you are working toward and excited to incorporate into Astra?

What are the most interesting, innovative, or unexpected ways that you have seen Astra used?

What are the most interesting, unexpected, or challenging lessons that you have learned while working on Astra?

When is Astra the wrong choice?

What do you have planned for the future of Astra?

Contact Info

Prabhat

LinkedIn @prabhatja on Twitter prabhatja on GitHub

Jonathan

LinkedIn @spyced on Twitter

Parting Question

From your perspective, what is the biggest gap in the tooling or technology for data management today?

Links

Pulsar

Podcast Episode Streamnative Episode

Datastax Astra Streaming Datastax Astra DB Luna Streaming Distribution Datastax Cassandra Kesque (formerly Kafkaesque) Kafka RabbitMQ Prometheus Grafana Pulsar Heartbeat Pulsar Summit Pulsar Summit Presentation on Kafka Connectors Replicated Chaos Engineering Fallout chaos engineering tools Jepsen

Podcast Episode

Jack VanLightly

BookKeeper TLA+ Model

Change Data Capture

The intro and outro music is from The Hug by The Freak Fandango Orchestra / CC BY-SA

Support Data Engineering Podcast

Summary Data quality is a concern that has been gaining attention alongside the rising importance of analytics for business success. Many solutions rely on hand-coded rules for catching known bugs, or statistical analysis of records to detect anomalies retroactively. While those are useful tools, it is far better to prevent data errors before they become an outsized issue. In this episode Gleb Mezhanskiy shares some strategies for adding quality checks at every stage of your development and deployment workflow to identify and fix problematic changes to your data before they get to production.

Announcements

Hello and welcome to the Data Engineering Podcast, the show about modern data management You listen to this show to learn about all of the latest tools, patterns, and practices that power data engineering projects across every domain. Now there’s a book that captures the foundational lessons and principles that underly everything that you hear about here. I’m happy to announce I collected wisdom from the community to help you in your journey as a data engineer and worked with O’Reilly to publish it as 97 Things Every Data Engineer Should Know. Go to dataengineeringpodcast.com/97things today to get your copy! When you’re ready to build your next pipeline, or want to test out the projects you hear about on the show, you’ll need somewhere to deploy it, so check out our friends at Linode. With their managed Kubernetes platform it’s now even easier to deploy and scale your workflows, or try out the latest Helm charts from tools like Pulsar and Pachyderm. With simple pricing, fast networking, object storage, and worldwide data centers, you’ve got everything you need to run a bulletproof data platform. Go to dataengineeringpodcast.com/linode today and get a $100 credit to try out a Kubernetes cluster of your own. And don’t forget to thank them for their continued support of this show! RudderStack’s smart customer data pipeline is warehouse-first. It builds your customer data warehouse and your identity graph on your data warehouse, with support for Snowflake, Google BigQuery, Amazon Redshift, and more. Their SDKs and plugins make event streaming easy, and their integrations with cloud applications like Salesforce and ZenDesk help you go beyond event streaming. With RudderStack you can use all of your customer data to answer more difficult questions and then send those insights to your whole customer data stack. Sign up free at dataengineeringpodcast.com/rudder today. We’ve all been asked to help with an ad-hoc request for data by the sales and marketing team. Then it becomes a critical report that they need updated every week or every day. Then what do you do? Send a CSV via email? Write some Python scripts to automate it? But what about incremental sync, API quotas, error handling, and all of the other details that eat up your time? Today, there is a better way. With Census, just write SQL or plug in your dbt models and start syncing your cloud warehouse to SaaS applications like Salesforce, Marketo, Hubspot, and many more. Go to dataengineeringpodcast.com/census today to get a free 14-day trial. Your host is Tobias Macey and today I’m interviewing Gleb Mezhanskiy about strategies for proactive data quality management and his work at Datafold to help provide tools for implementing them

Interview

Introduction How did you get involved in the area of data management? Can you describe what you are building at Datafold and the story behind it? What are the biggest factors that you see contributing to data quality issues?

How are teams identifying and addressing those failures?

How does the data platform architecture impact the potential for introducing quality problems? What are some of the potential risks or consequences of introducing errors in data processing? How can organizations shift to being proactive in their data quality management?

How much of a role does tooling play in addressing the introduct

Summary We have been building platforms and workflows to store, process, and analyze data since the earliest days of computing. Over that time there have been countless architectures, patterns, and "best practices" to make that task manageable. With the growing popularity of cloud services a new pattern has emerged and been dubbed the "Modern Data Stack". In this episode members of the GoDataDriven team, Guillermo Sanchez, Bram Ochsendorf, and Juan Perafan, explain the combinations of services that comprise this architecture, share their experiences working with clients to employ the stack, and the benefits of bringing engineers and business users together with data.

Announcements

Hello and welcome to the Data Engineering Podcast, the show about modern data management You listen to this show to learn about all of the latest tools, patterns, and practices that power data engineering projects across every domain. Now there’s a book that captures the foundational lessons and principles that underly everything that you hear about here. I’m happy to announce I collected wisdom from the community to help you in your journey as a data engineer and worked with O’Reilly to publish it as 97 Things Every Data Engineer Should Know. Go to dataengineeringpodcast.com/97things today to get your copy! When you’re ready to build your next pipeline, or want to test out the projects you hear about on the show, you’ll need somewhere to deploy it, so check out our friends at Linode. With their managed Kubernetes platform it’s now even easier to deploy and scale your workflows, or try out the latest Helm charts from tools like Pulsar and Pachyderm. With simple pricing, fast networking, object storage, and worldwide data centers, you’ve got everything you need to run a bulletproof data platform. Go to dataengineeringpodcast.com/linode today and get a $100 credit to try out a Kubernetes cluster of your own. And don’t forget to thank them for their continued support of this show! RudderStack’s smart customer data pipeline is warehouse-first. It builds your customer data warehouse and your identity graph on your data warehouse, with support for Snowflake, Google BigQuery, Amazon Redshift, and more. Their SDKs and plugins make event streaming easy, and their integrations with cloud applications like Salesforce and ZenDesk help you go beyond event streaming. With RudderStack you can use all of your customer data to answer more difficult questions and then send those insights to your whole customer data stack. Sign up free at dataengineeringpodcast.com/rudder today. We’ve all been asked to help with an ad-hoc request for data by the sales and marketing team. Then it becomes a critical report that they need updated every week or every day. Then what do you do? Send a CSV via email? Write some Python scripts to automate it? But what about incremental sync, API quotas, error handling, and all of the other details that eat up your time? Today, there is a better way. With Census, just write SQL or plug in your dbt models and start syncing your cloud warehouse to SaaS applications like Salesforce, Marketo, Hubspot, and many more. Go to dataengineeringpodcast.com/census today to get a free 14-day trial. Your host is Tobias Macey and today I’m interviewing Guillermo Sanchez, Bram Ochsendorf, and Juan Perafan about their experiences with managed services in the modern data stack in their work as consultants at GoDataDriven

Interview

Introduction How did you get involved in the area of data management? Can you start by giving your definition of the modern data stack?

What are the key characteristics of a tool or platform that make it a candidate for the "modern" stack?

How does the modern data stack shift the responsibilities and capabilities of data professionals and consumers? What are some difficulties that you face when working with customers to migrate to these new architectures? What are some of the limitations of the components or

Summary At the core of every data pipeline is an workflow manager (or several). Deploying, managing, and scaling that orchestration can consume a large fraction of a data team’s energy so it is important to pick something that provides the power and flexibility that you need. SaaSGlue is a managed service that lets you connect all of your systems, across clouds and physical infrastructure, and spanning all of your programming languages. In this episode Bart and Rich Wood explain how SaaSGlue is architected to allow for a high degree of flexibility in usage and deployment, their experience building a business with family, and how you can get started using it today. This is a fascinating platform with an endless set of use cases and a great team of people behind it.

Announcements

Hello and welcome to the Data Engineering Podcast, the show about modern data management When you’re ready to build your next pipeline, or want to test out the projects you hear about on the show, you’ll need somewhere to deploy it, so check out our friends at Linode. With their managed Kubernetes platform it’s now even easier to deploy and scale your workflows, or try out the latest Helm charts from tools like Pulsar and Pachyderm. With simple pricing, fast networking, object storage, and worldwide data centers, you’ve got everything you need to run a bulletproof data platform. Go to dataengineeringpodcast.com/linode today and get a $100 credit to try out a Kubernetes cluster of your own. And don’t forget to thank them for their continued support of this show! RudderStack’s smart customer data pipeline is warehouse-first. It builds your customer data warehouse and your identity graph on your data warehouse, with support for Snowflake, Google BigQuery, Amazon Redshift, and more. Their SDKs and plugins make event streaming easy, and their integrations with cloud applications like Salesforce and ZenDesk help you go beyond event streaming. With RudderStack you can use all of your customer data to answer more difficult questions and then send those insights to your whole customer data stack. Sign up free at dataengineeringpodcast.com/rudder today. We’ve all been asked to help with an ad-hoc request for data by the sales and marketing team. Then it becomes a critical report that they need updated every week or every day. Then what do you do? Send a CSV via email? Write some Python scripts to automate it? But what about incremental sync, API quotas, error handling, and all of the other details that eat up your time? Today, there is a better way. With Census, just write SQL or plug in your dbt models and start syncing your cloud warehouse to SaaS applications like Salesforce, Marketo, Hubspot, and many more. Go to dataengineeringpodcast.com/census today to get a free 14-day trial. Your host is Tobias Macey and today I’m interviewing Rich and Bart Wood about SaasGlue, a SaaS-based integration, orchestration and automation platform that lets you fill the gaps in your existing automation infrastructure

Interview

Introduction How did you get involved in the area of data management? Can you describe what SaasGlue is and the story behind it?

I understand that you are building this company with your 3 brothers. What have been the pros and cons of working with your family on this project?

What are the main use cases that you are focused on enabling?

Who are your target users and how has that influenced the features and design of the platform?

Orchestration, automation, and workflow management are all areas that have a range of active products and projects. How do you characterize SaaSGlue’s position in the overall ecosystem?

What are some of the ways that you see it integrated into a data platform?

What are the core elements and concepts of the SaaSGlue platform? How is the SaaSGlue platform architected?

How have the goals and design of the platform changed or evolved since you first began working on it? What are some of the assumptio

Summary While the overall concept of timeseries data is uniform, its usage and applications are far from it. One of the most demanding applications of timeseries data is for application and server monitoring due to the problem of high cardinality. In his quest to build a generalized platform for managing timeseries Paul Dix keeps getting pulled back into the monitoring arena. In this episode he shares the history of the InfluxDB project, the business that he has helped to build around it, and the architectural aspects of the engine that allow for its flexibility in managing various forms of timeseries data. This is a fascinating exploration of the technical and organizational evolution of the Influx Data platform, with some promising glimpses of where they are headed in the near future.

Announcements

Hello and welcome to the Data Engineering Podcast, the show about modern data management When you’re ready to build your next pipeline, or want to test out the projects you hear about on the show, you’ll need somewhere to deploy it, so check out our friends at Linode. With their managed Kubernetes platform it’s now even easier to deploy and scale your workflows, or try out the latest Helm charts from tools like Pulsar and Pachyderm. With simple pricing, fast networking, object storage, and worldwide data centers, you’ve got everything you need to run a bulletproof data platform. Go to dataengineeringpodcast.com/linode today and get a $100 credit to try out a Kubernetes cluster of your own. And don’t forget to thank them for their continued support of this show! RudderStack’s smart customer data pipeline is warehouse-first. It builds your customer data warehouse and your identity graph on your data warehouse, with support for Snowflake, Google BigQuery, Amazon Redshift, and more. Their SDKs and plugins make event streaming easy, and their integrations with cloud applications like Salesforce and ZenDesk help you go beyond event streaming. With RudderStack you can use all of your customer data to answer more difficult questions and then send those insights to your whole customer data stack. Sign up free at dataengineeringpodcast.com/rudder today. We’ve all been asked to help with an ad-hoc request for data by the sales and marketing team. Then it becomes a critical report that they need updated every week or every day. Then what do you do? Send a CSV via email? Write some Python scripts to automate it? But what about incremental sync, API quotas, error handling, and all of the other details that eat up your time? Today, there is a better way. With Census, just write SQL or plug in your dbt models and start syncing your cloud warehouse to SaaS applications like Salesforce, Marketo, Hubspot, and many more. Go to dataengineeringpodcast.com/census today to get a free 14-day trial. Your host is Tobias Macey and today I’m interviewing Paul Dix about Influx Data and the different facets of the market for timeseries databases

Interview

Introduction How did you get involved in the area of data management? Can you describe what you are building at Influx Data and the story behind it? Timeseries data is a fairly broad category with many variations in terms of storage volume, frequency, processing requirements, etc. This has led to an explosion of database engines and related tools to address these different needs. How do you think about your position and role in the ecosystem?

Who are your target customers and how does that focus inform your product and feature priorities? What are the use cases that Influx is best suited for?

Can you give an overview of the different projects, tools, and services that comprise your platform? How is InfluxDB architected?

How have the design and implementation of the DB engine changed or evolved since you first began working on it? What are you optimizing for on the consistency vs. availability spectrum of CAP? What is your approach to clustering/data distribution beyond a single node?

Summary The database is the core of any system because it holds the data that drives your entire experience. We spend countless hours designing the data model, updating engine versions, and tuning performance. But how confident are you that you have configured it to be as performant as possible, given the dozens of parameters and how they interact with each other? Andy Pavlo researches autonomous database systems, and out of that research he created OtterTune to find the optimal set of parameters to use for your specific workload. In this episode he explains how the system works, the challenge of scaling it to work across different database engines, and his hopes for the future of database systems.

Announcements

Hello and welcome to the Data Engineering Podcast, the show about modern data management When you’re ready to build your next pipeline, or want to test out the projects you hear about on the show, you’ll need somewhere to deploy it, so check out our friends at Linode. With their managed Kubernetes platform it’s now even easier to deploy and scale your workflows, or try out the latest Helm charts from tools like Pulsar and Pachyderm. With simple pricing, fast networking, object storage, and worldwide data centers, you’ve got everything you need to run a bulletproof data platform. Go to dataengineeringpodcast.com/linode today and get a $100 credit to try out a Kubernetes cluster of your own. And don’t forget to thank them for their continued support of this show! RudderStack’s smart customer data pipeline is warehouse-first. It builds your customer data warehouse and your identity graph on your data warehouse, with support for Snowflake, Google BigQuery, Amazon Redshift, and more. Their SDKs and plugins make event streaming easy, and their integrations with cloud applications like Salesforce and ZenDesk help you go beyond event streaming. With RudderStack you can use all of your customer data to answer more difficult questions and then send those insights to your whole customer data stack. Sign up free at dataengineeringpodcast.com/rudder today. We’ve all been asked to help with an ad-hoc request for data by the sales and marketing team. Then it becomes a critical report that they need updated every week or every day. Then what do you do? Send a CSV via email? Write some Python scripts to automate it? But what about incremental sync, API quotas, error handling, and all of the other details that eat up your time? Today, there is a better way. With Census, just write SQL or plug in your dbt models and start syncing your cloud warehouse to SaaS applications like Salesforce, Marketo, Hubspot, and many more. Go to dataengineeringpodcast.com/census today to get a free 14-day trial. Your host is Tobias Macey and today I’m interviewing Andy Pavlo about OtterTune, a system to continuously monitor and improve database performance via machine learning

Interview

Introduction How did you get involved in the area of data management? Can you describe what OtterTune is and the story behind it?

How does it relate to your work with NoisePage?

What are the challenges that database administrators, operators, and users run into when working with, configuring, and tuning transactional systems?

What are some of the contributing factors to the sprawling complexity of the configurable parameters for these databases?

Can you describe how OtterTune is implemented?

What are some of the aggregate benefits that OtterTune can gain by running as a centralized service and learning from all of the systems that it connects to? What are some of the assumptions that you made when starting the commercialization of this technology that have been challenged or invalidated as you began working with initial customers? How have the design and goals of the system changed or evolved since you first began working on it?

What is involved in adding support for a new database engine?

How applicable are the OtterTune capabilities to analyti

Summary When you build a machine learning model, the first step is always to load your data. Typically this means downloading files from object storage, or querying a database. To speed up the process, why not build the model inside the database so that you don’t have to move the information? In this episode Paige Roberts explains the benefits of pushing the machine learning processing into the database layer and the approach that Vertica has taken for their implementation. If you are looking for a way to speed up your experimentation, or an easy way to apply AutoML then this conversation is for you.

Announcements

Hello and welcome to the Data Engineering Podcast, the show about modern data management When you’re ready to build your next pipeline, or want to test out the projects you hear about on the show, you’ll need somewhere to deploy it, so check out our friends at Linode. With their managed Kubernetes platform it’s now even easier to deploy and scale your workflows, or try out the latest Helm charts from tools like Pulsar and Pachyderm. With simple pricing, fast networking, object storage, and worldwide data centers, you’ve got everything you need to run a bulletproof data platform. Go to dataengineeringpodcast.com/linode today and get a $100 credit to try out a Kubernetes cluster of your own. And don’t forget to thank them for their continued support of this show! RudderStack’s smart customer data pipeline is warehouse-first. It builds your customer data warehouse and your identity graph on your data warehouse, with support for Snowflake, Google BigQuery, Amazon Redshift, and more. Their SDKs and plugins make event streaming easy, and their integrations with cloud applications like Salesforce and ZenDesk help you go beyond event streaming. With RudderStack you can use all of your customer data to answer more difficult questions and then send those insights to your whole customer data stack. Sign up free at dataengineeringpodcast.com/rudder today. We’ve all been asked to help with an ad-hoc request for data by the sales and marketing team. Then it becomes a critical report that they need updated every week or every day. Then what do you do? Send a CSV via email? Write some Python scripts to automate it? But what about incremental sync, API quotas, error handling, and all of the other details that eat up your time? Today, there is a better way. With Census, just write SQL or plug in your dbt models and start syncing your cloud warehouse to SaaS applications like Salesforce, Marketo, Hubspot, and many more. Go to dataengineeringpodcast.com/census today to get a free 14-day trial. Your host is Tobias Macey and today I’m interviewing Paige Roberts about machine learning workflows inside the database

Interview

Introduction How did you get involved in the area of data management? Can you start by giving an overview of the current state of the market for databases that support in-process machine learning?

What are the motivating factors for running a machine learning workflow inside the database?

What styles of ML are feasible to do inside the database? (e.g. bayesian inference, deep learning, etc.) What are the performance implications of running a model training pipeline within the database runtime? (both in terms of training performance boosts, and database performance impacts) Can you describe the architecture of how the machine learning process is managed by the database engine? How do you manage interacting with Python/R/Jupyter/etc. when working within the database? What is the impact on data pipeline and MLOps architectures when using the database to manage the machine learning workflow? What are the most interesting, innovative, or unexpected ways that you have seen in-database ML used? What are the most interesting, unexpected, or challenging lessons that you have learned while working on machine learning inside the database? When is in-database ML the wrong choice? What are the recent trends/

Summary The way to build maintainable software and systems is through composition of individual pieces. By making those pieces high quality and flexible they can be used in surprising ways that the original creators couldn’t have imagined. One such component that has gone above and beyond its originally envisioned use case is BookKeeper, a distributed storage system that is optimized for durability and speed. In this episode Matteo Merli shares the story behind the creation of BookKeeper, the various ways that it is being used today, and the architectural aspects that make it such a strong building block for projects such as Pulsar. He also shares some of the other interesting systems that have been built on top of it and an amusing war story of running it at scale in its early years.

Announcements

Hello and welcome to the Data Engineering Podcast, the show about modern data management When you’re ready to build your next pipeline, or want to test out the projects you hear about on the show, you’ll need somewhere to deploy it, so check out our friends at Linode. With their managed Kubernetes platform it’s now even easier to deploy and scale your workflows, or try out the latest Helm charts from tools like Pulsar and Pachyderm. With simple pricing, fast networking, object storage, and worldwide data centers, you’ve got everything you need to run a bulletproof data platform. Go to dataengineeringpodcast.com/linode today and get a $100 credit to try out a Kubernetes cluster of your own. And don’t forget to thank them for their continued support of this show! RudderStack’s smart customer data pipeline is warehouse-first. It builds your customer data warehouse and your identity graph on your data warehouse, with support for Snowflake, Google BigQuery, Amazon Redshift, and more. Their SDKs and plugins make event streaming easy, and their integrations with cloud applications like Salesforce and ZenDesk help you go beyond event streaming. With RudderStack you can use all of your customer data to answer more difficult questions and then send those insights to your whole customer data stack. Sign up free at dataengineeringpodcast.com/rudder today. We’ve all been asked to help with an ad-hoc request for data by the sales and marketing team. Then it becomes a critical report that they need updated every week or every day. Then what do you do? Send a CSV via email? Write some Python scripts to automate it? But what about incremental sync, API quotas, error handling, and all of the other details that eat up your time? Today, there is a better way. With Census, just write SQL or plug in your dbt models and start syncing your cloud warehouse to SaaS applications like Salesforce, Marketo, Hubspot, and many more. Go to dataengineeringpodcast.com/census today to get a free 14-day trial. Your host is Tobias Macey and today I’m interviewing Matteo Merli about Apache BookKeeper, a scalable, fault-tolerant, and low-latency storage service optimized for real-time workloads

Interview

Introduction How did you get involved in the area of data management? Can you describe what BookKeeper is and the story behind it? What are the most notable features/capabilities of BookKeeper? What are some of the ways that BookKeeper is being used? How has your work on Pulsar influenced the features and product direction of BookKeeper? Can you describe the architecture of a BookKeeper cluster?

How have the design and goals of BookKeeper changed or evolved over time?

What is the impact of record-oriented storage on data distribution/allocation within the cluster when working with variable record sizes? What are some of the operational considerations that users should be aware of? What are some of the most interesting/compelling features from your perspective? What are some of the most often overlooked or misunderstood capabilities of BookKeeper? What are the most interesting, innovative, or unexpected ways that you have seen BookKeeper used? What

Summary The data warehouse has become the central component of the modern data stack. Building on this pattern, the team at Hightouch have created a platform that synchronizes information about your customers out to third party systems for use by marketing and sales teams. In this episode Tejas Manohar explains the benefits of sourcing customer data from one location for all of your organization to use, the technical challenges of synchronizing the data to external systems with varying APIs, and the workflow for enabling self-service access to your customer data by your marketing teams. This is an interesting conversation about the importance of the data warehouse and how it can be used beyond just internal analytics.

Announcements

Hello and welcome to the Data Engineering Podcast, the show about modern data management When you’re ready to build your next pipeline, or want to test out the projects you hear about on the show, you’ll need somewhere to deploy it, so check out our friends at Linode. With their managed Kubernetes platform it’s now even easier to deploy and scale your workflows, or try out the latest Helm charts from tools like Pulsar and Pachyderm. With simple pricing, fast networking, object storage, and worldwide data centers, you’ve got everything you need to run a bulletproof data platform. Go to dataengineeringpodcast.com/linode today and get a $60 credit to try out a Kubernetes cluster of your own. And don’t forget to thank them for their continued support of this show! Modern Data teams are dealing with a lot of complexity in their data pipelines and analytical code. Monitoring data quality, tracing incidents, and testing changes can be daunting and often takes hours to days. Datafold helps Data teams gain visibility and confidence in the quality of their analytical data through data profiling, column-level lineage and intelligent anomaly detection. Datafold also helps automate regression testing of ETL code with its Data Diff feature that instantly shows how a change in ETL or BI code affects the produced data, both on a statistical level and down to individual rows and values. Datafold integrates with all major data warehouses as well as frameworks such as Airflow & dbt and seamlessly plugs into CI workflows. Go to dataengineeringpodcast.com/datafold today to start a 30-day trial of Datafold. Once you sign up and create an alert in Datafold for your company data, they will send you a cool water flask. This episode of Data Engineering Podcast is sponsored by Datadog, a unified monitoring and analytics platform built for developers, IT operations teams, and businesses in the cloud age. Datadog provides customizable dashboards, log management, and machine-learning-based alerts in one fully-integrated platform so you can seamlessly navigate, pinpoint, and resolve performance issues in context. Monitor all your databases, cloud services, containers, and serverless functions in one place with Datadog’s 400+ vendor-backed integrations. If an outage occurs, Datadog provides seamless navigation between your logs, infrastructure metrics, and application traces in just a few clicks to minimize downtime. Try it yourself today by starting a free 14-day trial and receive a Datadog t-shirt after installing the agent. Go to dataengineeringpodcast.com/datadog today to see how you can enhance visibility into your stack with Datadog. Your host is Tobias Macey and today I’m interviewing Tejas Manohar about Hightouch, a data platform that helps you sync your customer data from your data warehouse to your CRM, marketing, and support tools

Interview

Introduction How did you get involved in the area of data management? Can you start by giving an overview of what you are building at Hightouch and your motivation for creating it? What are the main points of friction for teams who are trying to make use of customer data? Where is Hightouch positioned in the ecosystem of customer data tools such as Segment, Mixpanel

Summary In order to scale the use of data across an organization there are a number of challenges related to discovery, governance, and integration that need to be solved. The key to those solutions is a robust and flexible metadata management system. LinkedIn has gone through several iterations on the most maintainable and scalable approach to metadata, leading them to their current work on DataHub. In this episode Mars Lan and Pardhu Gunnam explain how they designed the platform, how it integrates into their data platforms, and how it is being used to power data discovery and analytics at LinkedIn.

Announcements

Hello and welcome to the Data Engineering Podcast, the show about modern data management What are the pieces of advice that you wish you had received early in your career of data engineering? If you hand a book to a new data engineer, what wisdom would you add to it? I’m working with O’Reilly on a project to collect the 97 things that every data engineer should know, and I need your help. Go to dataengineeringpodcast.com/97things to add your voice and share your hard-earned expertise. When you’re ready to build your next pipeline, or want to test out the projects you hear about on the show, you’ll need somewhere to deploy it, so check out our friends at Linode. With their managed Kubernetes platform it’s now even easier to deploy and scale your workflows, or try out the latest Helm charts from tools like Pulsar and Pachyderm. With simple pricing, fast networking, object storage, and worldwide data centers, you’ve got everything you need to run a bulletproof data platform. Go to dataengineeringpodcast.com/linode today and get a $60 credit to try out a Kubernetes cluster of your own. And don’t forget to thank them for their continued support of this show! If you’ve been exploring scalable, cost-effective and secure ways to collect and route data across your organization, RudderStack is the only solution that helps you turn your own warehouse into a state of the art customer data platform. Their mission is to empower data engineers to fully own their customer data infrastructure and easily push value to other parts of the organization, like marketing and product management. With their open-source foundation, fixed pricing, and unlimited volume, they are enterprise ready, but accessible to everyone. Go to dataengineeringpodcast.com/rudder to request a demo and get one free month of access to the hosted platform along with a free t-shirt. You listen to this show to learn and stay up to date with what’s happening in databases, streaming platforms, big data, and everything else you need to know about modern data platforms. For more opportunities to stay up to date, gain new skills, and learn from your peers there are a growing number of virtual events that you can attend from the comfort and safety of your home. Go to dataengineeringpodcast.com/conferences to check out the upcoming events being offered by our partners and get registered today! Your host is Tobias Macey and today I’m interviewing Pardhu Gunnam and Mars Lan about DataHub, LinkedIn’s metadata management and data catalog platform

Interview

Introduction How did you get involved in the area of data management? Can you start by giving an overview of what DataHub is and some of its back story?

What were you using at LinkedIn for metadata management prior to the introduction of DataHub? What was lacking in the previous solutions that motivated you to create a new platform?

There are a large number of other systems available for building data catalogs and tracking metadata, both open source and proprietary. What are the features of DataHub that would lead someone to use it in place of the other options? Who is the target audience for DataHub?

How do the needs of those end users influence or constrain your approach to the design and interfaces provided by DataHub?

Can you describe how DataHub is architected?

How has it evolved since yo

Summary Event based data is a rich source of information for analytics, unless none of the event structures are consistent. The team at Iteratively are building a platform to manage the end to end flow of collaboration around what events are needed, how to structure the attributes, and how they are captured. In this episode founders Patrick Thompson and Ondrej Hrebicek discuss the problems that they have experienced as a result of inconsistent event schemas, how the Iteratively platform integrates the definition, development, and delivery of event data, and the benefits of elevating the visibility of event data for improving the effectiveness of the resulting analytics. If you are struggling with inconsistent implementations of event data collection, lack of clarity on what attributes are needed, and how it is being used then this is definitely a conversation worth following.

Announcements

Hello and welcome to the Data Engineering Podcast, the show about modern data management What are the pieces of advice that you wish you had received early in your career of data engineering? If you hand a book to a new data engineer, what wisdom would you add to it? I’m working with O’Reilly on a project to collect the 97 things that every data engineer should know, and I need your help. Go to dataengineeringpodcast.com/97things to add your voice and share your hard-earned expertise. When you’re ready to build your next pipeline, or want to test out the projects you hear about on the show, you’ll need somewhere to deploy it, so check out our friends at Linode. With their managed Kubernetes platform it’s now even easier to deploy and scale your workflows, or try out the latest Helm charts from tools like Pulsar and Pachyderm. With simple pricing, fast networking, object storage, and worldwide data centers, you’ve got everything you need to run a bulletproof data platform. Go to dataengineeringpodcast.com/linode today and get a $60 credit to try out a Kubernetes cluster of your own. And don’t forget to thank them for their continued support of this show! If you’ve been exploring scalable, cost-effective and secure ways to collect and route data across your organization, RudderStack is the only solution that helps you turn your own warehouse into a state of the art customer data platform. Their mission is to empower data engineers to fully own their customer data infrastructure and easily push value to other parts of the organization, like marketing and product management. With their open-source foundation, fixed pricing, and unlimited volume, they are enterprise ready, but accessible to everyone. Go to dataengineeringpodcast.com/rudder to request a demo and get one free month of access to the hosted platform along with a free t-shirt. You listen to this show to learn and stay up to date with what’s happening in databases, streaming platforms, big data, and everything else you need to know about modern data platforms. For more opportunities to stay up to date, gain new skills, and learn from your peers there are a growing number of virtual events that you can attend from the comfort and safety of your home. Go to dataengineeringpodcast.com/conferences to check out the upcoming events being offered by our partners and get registered today! Your host is Tobias Macey and today I’m interviewing Patrick Thompson and Ondrej Hrebicek about Iteratively, a platform for enforcing consistent schemas for your event data

Interview

Introduction How did you get involved in the area of data management? Can you start by describing what you are building at Iteratively and your motivation for creating it? What are some of the ways that you have seen inconsistent message structures cause problems? What are some of the common anti-patterns that you have seen for managing the structure of event messages? What are the benefits that Iteratively provides for the different roles in an organization? Can you describe the workflow for a team using

Summary A majority of the scalable data processing platforms that we rely on are built as distributed systems. This brings with it a vast number of subtle ways that errors can creep in. Kyle Kingsbury created the Jepsen framework for testing the guarantees of distributed data processing systems and identifying when and why they break. In this episode he shares his approach to testing complex systems, the common challenges that are faced by engineers who build them, and why it is important to understand their limitations. This was a great look at some of the underlying principles that power your mission critical workloads.

Announcements

Hello and welcome to the Data Engineering Podcast, the show about modern data management What are the pieces of advice that you wish you had received early in your career of data engineering? If you hand a book to a new data engineer, what wisdom would you add to it? I’m working with O’Reilly on a project to collect the 97 things that every data engineer should know, and I need your help. Go to dataengineeringpodcast.com/97things to add your voice and share your hard-earned expertise. When you’re ready to build your next pipeline, or want to test out the projects you hear about on the show, you’ll need somewhere to deploy it, so check out our friends at Linode. With their managed Kubernetes platform it’s now even easier to deploy and scale your workflows, or try out the latest Helm charts from tools like Pulsar and Pachyderm. With simple pricing, fast networking, object storage, and worldwide data centers, you’ve got everything you need to run a bulletproof data platform. Go to dataengineeringpodcast.com/linode today and get a $60 credit to try out a Kubernetes cluster of your own. And don’t forget to thank them for their continued support of this show! If you’ve been exploring scalable, cost-effective and secure ways to collect and route data across your organization, RudderStack is the only solution that helps you turn your own warehouse into a state of the art customer data platform. Their mission is to empower data engineers to fully own their customer data infrastructure and easily push value to other parts of the organization, like marketing and product management. With their open-source foundation, fixed pricing, and unlimited volume, they are enterprise ready, but accessible to everyone. Go to dataengineeringpodcast.com/rudder to request a demo and get one free month of access to the hosted platform along with a free t-shirt. You listen to this show to learn and stay up to date with what’s happening in databases, streaming platforms, big data, and everything else you need to know about modern data platforms. For more opportunities to stay up to date, gain new skills, and learn from your peers there are a growing number of virtual events that you can attend from the comfort and safety of your home. Go to dataengineeringpodcast.com/conferences to check out the upcoming events being offered by our partners and get registered today! Your host is Tobias Macey and today I’m interviewing Kyle Kingsbury about his work on the Jepsen testing framework and the failure modes of distributed systems

Interview

Introduction How did you get involved in the area of data management? Can you start by describing what the Jepsen project is?

What was your inspiration for starting the project?

What other methods are available for evaluating and stress testing distributed systems? What are some of the common misconceptions or misunderstanding of distributed systems guarantees and how they impact real world usage of things like databases? How do you approach the design of a test suite for a new distributed system?

What is your heuristic for determining the completeness of your test suite?

What are some of the common challenges of setting up a representative deployment for testing? Can you walk through the workflow of setting up, running, and evaluating the output of a Jepsen test? Ho

Summary Object storage is quickly becoming the unifying layer for data intensive applications and analytics. Modern, cloud oriented data warehouses and data lakes both rely on the durability and ease of use that it provides. S3 from Amazon has quickly become the de-facto API for interacting with this service, so the team at MinIO have built a production grade, easy to manage storage engine that replicates that interface. In this episode Anand Babu Periasamy shares the origin story for the MinIO platform, the myriad use cases that it supports, and the challenges that they have faced in replicating the functionality of S3. He also explains the technical implementation, innovative design, and broad vision for the project.

Announcements

Hello and welcome to the Data Engineering Podcast, the show about modern data management When you’re ready to build your next pipeline, or want to test out the projects you hear about on the show, you’ll need somewhere to deploy it, so check out our friends at Linode. With 200Gbit private networking, scalable shared block storage, and a 40Gbit public network, you’ve got everything you need to run a fast, reliable, and bullet-proof data platform. If you need global distribution, they’ve got that covered too with world-wide datacenters including new ones in Toronto and Mumbai. And for your machine learning workloads, they just announced dedicated CPU instances. Go to dataengineeringpodcast.com/linode today to get a $20 credit and launch a new server in under a minute. And don’t forget to thank them for their continued support of this show! You listen to this show to learn and stay up to date with what’s happening in databases, streaming platforms, big data, and everything else you need to know about modern data management.For even more opportunities to meet, listen, and learn from your peers you don’t want to miss out on this year’s conference season. We have partnered with organizations such as O’Reilly Media, Dataversity, Corinium Global Intelligence, and Data Council. Upcoming events include the O’Reilly AI conference, the Strata Data conference, the combined events of the Data Architecture Summit and Graphorum, and Data Council in Barcelona. Go to dataengineeringpodcast.com/conferences to learn more about these and other events, and take advantage of our partner discounts to save money when you register today. Your host is Tobias Macey and today I’m interviewing Anand Babu Periasamy about MinIO, the neutral, open source, enterprise grade object storage system.

Interview

Introduction How did you get involved in the area of data management? Can you explain what MinIO is and its origin story? What are some of the main use cases that MinIO enables? How does MinIO compare to other object storage options and what benefits does it provide over other open source platforms?

Your marketing focuses on the utility of MinIO for ML and AI workloads. What benefits does object storage provide as compared to distributed file systems? (e.g. HDFS, GlusterFS, Ceph)

What are some of the challenges that you face in terms of maintaining compatibility with the S3 interface?

What are the constraints and opportunities that are provided by adhering to that API?

Can you describe how MinIO is implemented and the overall system design?

How has that design evolved since you first began working on it?

What assumptions did you have at the outset and how have they been challenged or updated?

What are the axes for scaling that MinIO provides and how does it handle clustering?

Where does it fall on the axes of availability and consistency in the CAP theorem?

One of the useful features that you provide is efficient erasure coding, as well as protection against data corruption. How much overhead do those capabilties incur, in terms of computational efficiency and, in a clustered scenario, storage volume? For someone who is interested in running MinIO, what is involved in deploying and maintain