ABOUT THE TALK As machine learning models are becoming more common in production, organizations are recognizing the significance of continuous validation, and are integrating automated testing into their CI/CD pipelines to ensure that their models remain relevant and are trustworthy. However, with constantly changing data and black-box logic, testing these models can be a daunting task.

In this talk, we explore the common pitfalls of ML models and best practices for testing them. We demonstrate how to use the deepchecks open source package to validate models and data during the research and CI/CD phases.

ABOUT THE SPEAKER Shir Chorev is the co-founder and CTO of Deepchecks, an MLOps startup for continuous validation of ML models and data. Previously, Shir worked at the Prime Minister’s Office and at Unit 8200, conducting and leading research in various Machine Learning and Cybersecurity related challenges.

ABOUT DATA COUNCIL: Data Council (https://www.datacouncil.ai/) is a community and conference series that provides data professionals with the learning and networking opportunities they need to grow their careers.

Make sure to subscribe to our channel for the most up-to-date talks from technical professionals on data related topics including data infrastructure, data engineering, ML systems, analytics and AI from top startups and tech companies.

FOLLOW DATA COUNCIL: Twitter: https://twitter.com/DataCouncilAI LinkedIn: https://www.linkedin.com/company/datacouncil-ai/

ABOUT THE TALK:
Our traditional testing and CI methods for Data Science are not working, but we can't just give up on providing guardrails.

As engineers, how do you solve ML testing?

In this talk, Emily Curtain discusses: - abstracting, decoupling, and separating concerns - keeping pytest only where it belongs - substituting testing for observability in appropriate places - applying data reliability practices and thereby solving some problems at the source - by honoring Data Scientists' mental models, and ways of working

ABOUT THE SPEAKER: Emily Curtin is a Staff MLOps Engineer at Intuit Mailchimp. She leads a crazy good team focused on helping Data Scientists do higher quality work faster and more intuitively.

ABOUT DATA COUNCIL: Data Council (https://www.datacouncil.ai/) is a community and conference series that provides data professionals with the learning and networking opportunities they need to grow their careers.

Make sure to subscribe to our channel for the most up-to-date talks from technical professionals on data related topics including data infrastructure, data engineering, ML systems, analytics and AI from top startups and tech companies.

FOLLOW DATA COUNCIL: Twitter: https://twitter.com/DataCouncilAI LinkedIn: https://www.linkedin.com/company/datacouncil-ai/

Developers often use code coverage as a target, which makes it a bad measure of test quality.

Mutation testing changes the game: create mutant versions of your code that break your tests, and you'll quickly start to write better tests!

Come and learn to use it as part of your CI/CD process. I promise, you'll never look at penguins the same way again!

Working with python is fun. Managing python packaging, linters, tests, CI, etc. is not as fun.

Every maintainer needs to worry about consistent styling, quality, speed of tests, etc as the project grows.

Monorepos have been successful in other communities - how does it work in Python ?

I will present the challenges we encountered while migrating an ML model from batch to real-time predictions and how we handled them. In particular, I will focus on the design decisions and open-source tools we built to test the code, data and models as part of the CI/CD pipeline and enable us to ship fast with confidence.

We will delve into the importance of effective data visualisation in today's world. We will explore how it can help convey insights from data using Matplotlib and best practices for creating informative visualisations. We will also discuss the limitations of static visualisations and examine the role of continuous integration in streamlining the process and avoiding common pitfalls. By the end of this talk, you will have gained valuable insights and techniques for creating informative and accurate data visualisations, no matter what tools you're using.

Summary

With all of the messaging about treating data as a product it is becoming difficult to know what that even means. Vishal Singh is the head of products at Starburst which means that he has to spend all of his time thinking and talking about the details of product thinking and its application to data. In this episode he shares his thoughts on the strategic and tactical elements of moving your work as a data professional from being task-oriented to being product-oriented and the long term improvements in your productivity that it provides.

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! 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 or even weeks. By the time errors have made their way into production, it’s often too late and damage is done. Datafold built automated regression testing to help data and analytics engineers deal with data quality in their pull requests. Datafold shows how a change in SQL code affects your data, both on a statistical level and down to individual rows and values before it gets merged to production. No more shipping and praying, you can now know exactly what will change in your database! Datafold integrates with all major data warehouses as well as frameworks such as Airflow & dbt and seamlessly plugs into CI workflows. Visit dataengineeringpodcast.com/datafold today to book a demo with Datafold. RudderStack helps you build a customer data platform on your warehouse or data lake. Instead of trapping data in a black box, they enable you to easily collect customer data from the entire stack and build an identity graph on your warehouse, giving you full visibility and control. Their SDKs make event streaming from any app or website easy, and their extensive library of integrations enable you to automatically send data to hundreds of downstream tools. Sign up free at dataengineeringpodcast.com/rudder Build Data Pipelines. Not DAGs. That’s the spirit behind Upsolver SQLake, a new self-service data pipeline platform that lets you build batch and streaming pipelines without falling into the black hole of DAG-based orchestration. All you do is write a query in SQL to declare your transformation, and SQLake will turn it into a continuous pipeline that scales to petabytes and delivers up to the minute fresh data. SQLake supports a broad set of transformations, including high-cardinality joins, aggregations, upserts and window operations. Output data can be streamed into a data lake for query engines like Presto, Trino or Spark SQL, a data warehouse like Snowflake or Redshift., or any other destination you choose. Pricing for SQLake is simple. You pay $99 per terabyte ingested into your data lake using SQLake, and run unlimited transformation pipelines for free. That way data engineers and data users can process to their heart’s content without worrying about their cloud bill. For data engineering podcast listeners, we’re offering a 30 day trial with unlimited data, so go to dataengineeringpodcast.com/upsolver today and see for yourself how to avoid DAG hell. Your host is Tobias Macey and today I'm interviewing Vishal Singh about his experience

Summary

Encryption and security are critical elements in data analytics and machine learning applications. We have well developed protocols and practices around data that is at rest and in motion, but security around data in use is still severely lacking. Recognizing this shortcoming and the capabilities that could be unlocked by a robust solution Rishabh Poddar helped to create Opaque Systems as an outgrowth of his PhD studies. In this episode he shares the work that he and his team have done to simplify integration of secure enclaves and trusted computing environments into analytical workflows and how you can start using it without re-engineering your existing 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 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! 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 or even weeks. By the time errors have made their way into production, it’s often too late and damage is done. Datafold built automated regression testing to help data and analytics engineers deal with data quality in their pull requests. Datafold shows how a change in SQL code affects your data, both on a statistical level and down to individual rows and values before it gets merged to production. No more shipping and praying, you can now know exactly what will change in your database! Datafold integrates with all major data warehouses as well as frameworks such as Airflow & dbt and seamlessly plugs into CI workflows. Visit dataengineeringpodcast.com/datafold today to book a demo with Datafold. RudderStack helps you build a customer data platform on your warehouse or data lake. Instead of trapping data in a black box, they enable you to easily collect customer data from the entire stack and build an identity graph on your warehouse, giving you full visibility and control. Their SDKs make event streaming from any app or website easy, and their extensive library of integrations enable you to automatically send data to hundreds of downstream tools. Sign up free at dataengineeringpodcast.com/rudder Build Data Pipelines. Not DAGs. That’s the spirit behind Upsolver SQLake, a new self-service data pipeline platform that lets you build batch and streaming pipelines without falling into the black hole of DAG-based orchestration. All you do is write a query in SQL to declare your transformation, and SQLake will turn it into a continuous pipeline that scales to petabytes and delivers up to the minute fresh data. SQLake supports a broad set of transformations, including high-cardinality joins, aggregations, upserts and window operations. Output data can be streamed into a data lake for query engines like Presto, Trino or Spark SQL, a data warehouse like Snowflake or Redshift., or any other destination you choose. Pricing for SQLake is simple. You pay $99 per terabyte ingested into your data lake using SQLake, and run unlimited transformation pipelines for free. That way data engineers and data users can process to their heart’s content without worrying about their cloud bill. For data engineering podcast listeners, we’re offering a 30 day trial with unlimited data, so go to dataengineeringpodcast.com/upsolver today an

Summary

The data ecosystem has seen a constant flurry of activity for the past several years, and it shows no signs of slowing down. With all of the products, techniques, and buzzwords being discussed it can be easy to be overcome by the hype. In this episode Juan Sequeda and Tim Gasper from data.world share their views on the core principles that you can use to ground your work and avoid getting caught in the hype cycles.

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! 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 or even weeks. By the time errors have made their way into production, it’s often too late and damage is done. Datafold built automated regression testing to help data and analytics engineers deal with data quality in their pull requests. Datafold shows how a change in SQL code affects your data, both on a statistical level and down to individual rows and values before it gets merged to production. No more shipping and praying, you can now know exactly what will change in your database! Datafold integrates with all major data warehouses as well as frameworks such as Airflow & dbt and seamlessly plugs into CI workflows. Visit dataengineeringpodcast.com/datafold today to book a demo with Datafold. RudderStack helps you build a customer data platform on your warehouse or data lake. Instead of trapping data in a black box, they enable you to easily collect customer data from the entire stack and build an identity graph on your warehouse, giving you full visibility and control. Their SDKs make event streaming from any app or website easy, and their extensive library of integrations enable you to automatically send data to hundreds of downstream tools. Sign up free at dataengineeringpodcast.com/rudder Build Data Pipelines. Not DAGs. That’s the spirit behind Upsolver SQLake, a new self-service data pipeline platform that lets you build batch and streaming pipelines without falling into the black hole of DAG-based orchestration. All you do is write a query in SQL to declare your transformation, and SQLake will turn it into a continuous pipeline that scales to petabytes and delivers up to the minute fresh data. SQLake supports a broad set of transformations, including high-cardinality joins, aggregations, upserts and window operations. Output data can be streamed into a data lake for query engines like Presto, Trino or Spark SQL, a data warehouse like Snowflake or Redshift., or any other destination you choose. Pricing for SQLake is simple. You pay $99 per terabyte ingested into your data lake using SQLake, and run unlimited transformation pipelines for free. That way data engineers and data users can process to their heart’s content without worrying about their cloud bill. For data engineering podcast listeners, we’re offering a 30 day trial with unlimited data, so go to dataengineeringpodcast.com/upsolver today and see for yourself how to avoid DAG hell. Your host is Tobias Macey and today I'm interviewing Juan Sequeda and Tim Gasper about their views on the role of the data mesh paradigm for driving re-assessment of the foundational principles of data systems

Summary One of the most critical aspects of software projects is managing its data. Managing the operational concerns for your database can be complex and expensive, especially if you need to scale to large volumes of data, high traffic, or geographically distributed usage. Planetscale is a serverless option for your MySQL workloads that lets you focus on your applications without having to worry about managing the database or fight with differences between development and production. In this episode Nick van Wiggeren explains how the Planetscale platform is implemented, their strategies for balancing maintenance and improvements of the underlying Vitess project with their business goals, and how you can start using it today to free up the time you spend on database administration.

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! 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 or even weeks. By the time errors have made their way into production, it’s often too late and damage is done. Datafold built automated regression testing to help data and analytics engineers deal with data quality in their pull requests. Datafold shows how a change in SQL code affects your data, both on a statistical level and down to individual rows and values before it gets merged to production. No more shipping and praying, you can now know exactly what will change in your database! Datafold integrates with all major data warehouses as well as frameworks such as Airflow & dbt and seamlessly plugs into CI workflows. Visit dataengineeringpodcast.com/datafold today to book a demo with Datafold. RudderStack helps you build a customer data platform on your warehouse or data lake. Instead of trapping data in a black box, they enable you to easily collect customer data from the entire stack and build an identity graph on your warehouse, giving you full visibility and control. Their SDKs make event streaming from any app or website easy, and their extensive library of integrations enable you to automatically send data to hundreds of downstream tools. Sign up free at dataengineeringpodcast.com/rudder Build Data Pipelines. Not DAGs. That’s the spirit behind Upsolver SQLake, a new self-service data pipeline platform that lets you build batch and streaming pipelines without falling into the black hole of DAG-based orchestration. All you do is write a query in SQL to declare your transformation, and SQLake will turn it into a continuous pipeline that scales to petabytes and delivers up to the minute fresh data. SQLake supports a broad set of transformations, including high-cardinality joins, aggregations, upserts and window operations. Output data can be streamed into a data lake for query engines like Presto, Trino or Spark SQL, a data warehouse like Snowflake or Redshift., or any other destination you choose. Pricing for SQLake is simple. You pay $99 per terabyte ingested into your data lake using SQLake, and run unlimited transformation pipelines for free. That way data engineers and data users can process to their heart’s content without worrying about their cloud bill. For data engineering podcast l

Summary The term "real-time data" brings with it a combination of excitement, uncertainty, and skepticism. The promise of insights that are always accurate and up to date is appealing to organizations, but the technical realities to make it possible have been complex and expensive. In this episode Arjun Narayan explains how the technical barriers to adopting real-time data in your analytics and applications have become surmountable by organizations of all sizes.

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! 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 or even weeks. By the time errors have made their way into production, it’s often too late and damage is done. Datafold built automated regression testing to help data and analytics engineers deal with data quality in their pull requests. Datafold shows how a change in SQL code affects your data, both on a statistical level and down to individual rows and values before it gets merged to production. No more shipping and praying, you can now know exactly what will change in your database! Datafold integrates with all major data warehouses as well as frameworks such as Airflow & dbt and seamlessly plugs into CI workflows. Visit dataengineeringpodcast.com/datafold today to book a demo with Datafold. RudderStack helps you build a customer data platform on your warehouse or data lake. Instead of trapping data in a black box, they enable you to easily collect customer data from the entire stack and build an identity graph on your warehouse, giving you full visibility and control. Their SDKs make event streaming from any app or website easy, and their extensive library of integrations enable you to automatically send data to hundreds of downstream tools. Sign up free at dataengineeringpodcast.com/rudder Build Data Pipelines. Not DAGs. That’s the spirit behind Upsolver SQLake, a new self-service data pipeline platform that lets you build batch and streaming pipelines without falling into the black hole of DAG-based orchestration. All you do is write a query in SQL to declare your transformation, and SQLake will turn it into a continuous pipeline that scales to petabytes and delivers up to the minute fresh data. SQLake supports a broad set of transformations, including high-cardinality joins, aggregations, upserts and window operations. Output data can be streamed into a data lake for query engines like Presto, Trino or Spark SQL, a data warehouse like Snowflake or Redshift., or any other destination you choose. Pricing for SQLake is simple. You pay $99 per terabyte ingested into your data lake using SQLake, and run unlimited transformation pipelines for free. That way data engineers and data users can process to their heart’s content without worrying about their cloud bill. For data engineering podcast listeners, we’re offering a 30 day trial with unlimited data, so go to dataengineeringpodcast.com/upsolver today and see for yourself how to avoid DAG hell. Your host is Tobias Macey and today I’m interviewing Arjun Narayan about the benefits of real-time data for teams of all sizes

Interview

Introduction How did you ge

Summary The most expensive part of working with massive data sets is the work of retrieving and processing the files that contain the raw information. FeatureBase (formerly Pilosa) avoids that overhead by converting the data into bitmaps. In this episode Matt Jaffee explains how to model your data as bitmaps and the benefits that this representation provides for fast aggregate computation. He also discusses the improvements that have been incorporated into FeatureBase to simplify integration with the rest of your data stack, and the SQL interface that was added to make working with the product easier.

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! 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 or even weeks. By the time errors have made their way into production, it’s often too late and damage is done. Datafold built automated regression testing to help data and analytics engineers deal with data quality in their pull requests. Datafold shows how a change in SQL code affects your data, both on a statistical level and down to individual rows and values before it gets merged to production. No more shipping and praying, you can now know exactly what will change in your database! Datafold integrates with all major data warehouses as well as frameworks such as Airflow & dbt and seamlessly plugs into CI workflows. Visit dataengineeringpodcast.com/datafold today to book a demo with Datafold. RudderStack helps you build a customer data platform on your warehouse or data lake. Instead of trapping data in a black box, they enable you to easily collect customer data from the entire stack and build an identity graph on your warehouse, giving you full visibility and control. Their SDKs make event streaming from any app or website easy, and their extensive library of integrations enable you to automatically send data to hundreds of downstream tools. Sign up free at dataengineeringpodcast.com/rudder Build Data Pipelines. Not DAGs. That’s the spirit behind Upsolver SQLake, a new self-service data pipeline platform that lets you build batch and streaming pipelines without falling into the black hole of DAG-based orchestration. All you do is write a query in SQL to declare your transformation, and SQLake will turn it into a continuous pipeline that scales to petabytes and delivers up to the minute fresh data. SQLake supports a broad set of transformations, including high-cardinality joins, aggregations, upserts and window operations. Output data can be streamed into a data lake for query engines like Presto, Trino or Spark SQL, a data warehouse like Snowflake or Redshift., or any other destination you choose. Pricing for SQLake is simple. You pay $99 per terabyte ingested into your data lake using SQLake, and run unlimited transformation pipelines for free. That way data engineers and data users can process to their heart’s content without worrying about their cloud bill. For data engineering podcast listeners, we’re offering a 30 day trial with unlimited data, so go to dataengineeringpodcast.com/upsolver today and see for yourself how to avoid DAG hell. Your host is Tobias Macey

Summary The problems that are easiest to fix are the ones that you prevent from happening in the first place. Sifflet is a platform that brings your entire data stack into focus to improve the reliability of your data assets and empower collaboration across your teams. In this episode CEO and founder Salma Bakouk shares her views on the causes and impacts of "data entropy" and how you can tame it before it leads to failures.

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! 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 or even weeks. By the time errors have made their way into production, it’s often too late and damage is done. Datafold built automated regression testing to help data and analytics engineers deal with data quality in their pull requests. Datafold shows how a change in SQL code affects your data, both on a statistical level and down to individual rows and values before it gets merged to production. No more shipping and praying, you can now know exactly what will change in your database! Datafold integrates with all major data warehouses as well as frameworks such as Airflow & dbt and seamlessly plugs into CI workflows. Visit dataengineeringpodcast.com/datafold today to book a demo with Datafold. RudderStack helps you build a customer data platform on your warehouse or data lake. Instead of trapping data in a black box, they enable you to easily collect customer data from the entire stack and build an identity graph on your warehouse, giving you full visibility and control. Their SDKs make event streaming from any app or website easy, and their extensive library of integrations enable you to automatically send data to hundreds of downstream tools. Sign up free at dataengineeringpodcast.com/rudder Data teams are increasingly under pressure to deliver. According to a recent survey by Ascend.io, 95% in fact reported being at or over capacity. With 72% of data experts reporting demands on their team going up faster than they can hire, it’s no surprise they are increasingly turning to automation. In fact, while only 3.5% report having current investments in automation, 85% of data teams plan on investing in automation in the next 12 months. 85%!!! That’s where our friends at Ascend.io come in. The Ascend Data Automation Cloud provides a unified platform for data ingestion, transformation, orchestration, and observability. Ascend users love its declarative pipelines, powerful SDK, elegant UI, and extensible plug-in architecture, as well as its support for Python, SQL, Scala, and Java. Ascend automates workloads on Snowflake, Databricks, BigQuery, and open source Spark, and can be deployed in AWS, Azure, or GCP. Go to dataengineeringpodcast.com/ascend and sign up for a free trial. If you’re a data engineering podcast listener, you get credits worth $5,000 when you become a customer. Your host is Tobias Macey and today I’m interviewing Salma Bakouk about achieving data reliability and reducing entropy within your data stack with sifflet

Interview

Introduction How did you get involved in the area of data management? Can you describe what Sifflet is and the st

Summary CreditKarma builds data products that help consumers take advantage of their credit and financial capabilities. To make that possible they need a reliable data platform that empowers all of the organization’s stakeholders. In this episode Vishnu Venkataraman shares the journey that he and his team have taken to build and evolve their systems and improve the product offerings that they are able to support.

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! 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 or even weeks. By the time errors have made their way into production, it’s often too late and damage is done. Datafold built automated regression testing to help data and analytics engineers deal with data quality in their pull requests. Datafold shows how a change in SQL code affects your data, both on a statistical level and down to individual rows and values before it gets merged to production. No more shipping and praying, you can now know exactly what will change in your database! Datafold integrates with all major data warehouses as well as frameworks such as Airflow & dbt and seamlessly plugs into CI workflows. Visit dataengineeringpodcast.com/datafold today to book a demo with Datafold. RudderStack helps you build a customer data platform on your warehouse or data lake. Instead of trapping data in a black box, they enable you to easily collect customer data from the entire stack and build an identity graph on your warehouse, giving you full visibility and control. Their SDKs make event streaming from any app or website easy, and their extensive library of integrations enable you to automatically send data to hundreds of downstream tools. Sign up free at dataengineeringpodcast.com/rudder Data teams are increasingly under pressure to deliver. According to a recent survey by Ascend.io, 95% in fact reported being at or over capacity. With 72% of data experts reporting demands on their team going up faster than they can hire, it’s no surprise they are increasingly turning to automation. In fact, while only 3.5% report having current investments in automation, 85% of data teams plan on investing in automation in the next 12 months. 85%!!! That’s where our friends at Ascend.io come in. The Ascend Data Automation Cloud provides a unified platform for data ingestion, transformation, orchestration, and observability. Ascend users love its declarative pipelines, powerful SDK, elegant UI, and extensible plug-in architecture, as well as its support for Python, SQL, Scala, and Java. Ascend automates workloads on Snowflake, Databricks, BigQuery, and open source Spark, and can be deployed in AWS, Azure, or GCP. Go to dataengineeringpodcast.com/ascend and sign up for a free trial. If you’re a data engineering podcast listener, you get credits worth $5,000 when you become a customer. Your host is Tobias Macey and today I’m interviewing Vishnu Venkataraman about building the data platform at CreditKarma and the forces that shaped the design

Interview

Introduction How did you get involved in the area of data management? Can you describe what CreditKarma is and the role

Summary Despite the best efforts of data engineers, data is as messy as the real world. Entity resolution and fuzzy matching are powerful utilities for cleaning up data from disconnected sources, but it has typically required custom development and training machine learning models. Sonal Goyal created and open-sourced Zingg as a generalized tool for data mastering and entity resolution to reduce the effort involved in adopting those practices. In this episode she shares the story behind the project, the details of how it is implemented, and how you can use it for your own data projects.

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! RudderStack helps you build a customer data platform on your warehouse or data lake. Instead of trapping data in a black box, they enable you to easily collect customer data from the entire stack and build an identity graph on your warehouse, giving you full visibility and control. Their SDKs make event streaming from any app or website easy, and their extensive library of integrations enable you to automatically send data to hundreds of downstream tools. Sign up free at dataengineeringpodcast.com/rudder 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 or even weeks. By the time errors have made their way into production, it’s often too late and damage is done. Datafold built automated regression testing to help data and analytics engineers deal with data quality in their pull requests. Datafold shows how a change in SQL code affects your data, both on a statistical level and down to individual rows and values before it gets merged to production. No more shipping and praying, you can now know exactly what will change in your database! Datafold integrates with all major data warehouses as well as frameworks such as Airflow & dbt and seamlessly plugs into CI workflows. Visit dataengineeringpodcast.com/datafold today to book a demo with Datafold. Data teams are increasingly under pressure to deliver. According to a recent survey by Ascend.io, 95% in fact reported being at or over capacity. With 72% of data experts reporting demands on their team going up faster than they can hire, it’s no surprise they are increasingly turning to automation. In fact, while only 3.5% report having current investments in automation, 85% of data teams plan on investing in automation in the next 12 months. 85%!!! That’s where our friends at Ascend.io come in. The Ascend Data Automation Cloud provides a unified platform for data ingestion, transformation, orchestration, and observability. Ascend users love its declarative pipelines, powerful SDK, elegant UI, and extensible plug-in architecture, as well as its support for Python, SQL, Scala, and Java. Ascend automates workloads on Snowflake, Databricks, BigQuery, and open source Spark, and can be deployed in AWS, Azure, or GCP. Go to dataengineeringpodcast.com/ascend and sign up for a free trial. If you’re a data engineering podcast listener, you get credits worth $5,000 when you become a customer. Your host is Tobias Macey and today I’m interviewing Sonal Goyal about Zingg, an open source entity resolution frame

Summary One of the most impactful technologies for data analytics in recent years has been dbt. It’s hard to have a conversation about data engineering or analysis without mentioning it. Despite its widespread adoption there are still rough edges in its workflow that cause friction for data analysts. To help simplify the adoption and management of dbt projects Nandam Karthik helped create Optimus. In this episode he shares his experiences working with organizations to adopt analytics engineering patterns and the ways that Optimus and dbt were combined to let data analysts deliver insights without the roadblocks of complex pipeline management.

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! 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 or even weeks. By the time errors have made their way into production, it’s often too late and damage is done. Datafold built automated regression testing to help data and analytics engineers deal with data quality in their pull requests. Datafold shows how a change in SQL code affects your data, both on a statistical level and down to individual rows and values before it gets merged to production. No more shipping and praying, you can now know exactly what will change in your database! Datafold integrates with all major data warehouses as well as frameworks such as Airflow & dbt and seamlessly plugs into CI workflows. Visit dataengineeringpodcast.com/datafold today to book a demo with Datafold. RudderStack helps you build a customer data platform on your warehouse or data lake. Instead of trapping data in a black box, they enable you to easily collect customer data from the entire stack and build an identity graph on your warehouse, giving you full visibility and control. Their SDKs make event streaming from any app or website easy, and their extensive library of integrations enable you to automatically send data to hundreds of downstream tools. Sign up free at dataengineeringpodcast.com/rudder Data teams are increasingly under pressure to deliver. According to a recent survey by Ascend.io, 95% in fact reported being at or over capacity. With 72% of data experts reporting demands on their team going up faster than they can hire, it’s no surprise they are increasingly turning to automation. In fact, while only 3.5% report having current investments in automation, 85% of data teams plan on investing in automation in the next 12 months. 85%!!! That’s where our friends at Ascend.io come in. The Ascend Data Automation Cloud provides a unified platform for data ingestion, transformation, orchestration, and observability. Ascend users love its declarative pipelines, powerful SDK, elegant UI, and extensible plug-in architecture, as well as its support for Python, SQL, Scala, and Java. Ascend automates workloads on Snowflake, Databricks, BigQuery, and open source Spark, and can be deployed in AWS, Azure, or GCP. Go to dataengineeringpodcast.com/ascend and sign up for a free trial. If you’re a data engineering podcast listener, you get credits worth $5,000 when you become a customer. Your host is Tobias Macey and today I’m interviewing Nand

Growing demand for data puts Analytics Engineers under pressure to move quickly, without introducing bugs to their models. If you find yourself pushing dbt’s testing framework to the limit, or waiting for careful code reviews before deploying to production, this talk is for you. Join Egor Gryaznov, CTO of Bigeye and former Staff Data Engineer at Uber, to learn how to take your workflow to the next level with blue-green deployments. This talk will cover: -CI/CD pipeline best practices for Analytics Engineers -Automatically validating pipeline changes with blue-green deployments -Tracking data changes through merge requests for visibility and reliability More details coming soon!

Check the slides here: https://docs.google.com/presentation/d/1zkzKkhjm1WshCMSOksfi24BEGK62wrLOGKKEg0I0Yhg/edit#slide=id.g11c331d5e28_0_6192

Coalesce 2023 is coming! Register for free at https://coalesce.getdbt.com/.

Does a two-step deployment workflow for developing, testing, and deploying code to dbt Cloud sound possible? Sunrun thinks so. Join James Sorensen and Jared Stout to learn how they used Github Actions and API integrations with dbt Cloud and Jira to entirely automate the CI/CD workflow, saving the team time and worry when moving through SOX certification.

Check the slides here: https://docs.google.com/presentation/d/1ZecU0-TN8SxNFpdKdkVksuDjpUy6XiaulqBdfqhLb68/edit#slide=id.g15507761f0b_0_10

Coalesce 2023 is coming! Register for free at https://coalesce.getdbt.com/.

Automation will make your dbt development faster and safer — but how do you get started? In this workshop we'll introduce you to the "Software Development Life Cycle" and some of the automation opportunities it offers from pre-commit to CI and beyond.

This workshop is ideal for the decision maker who is planning for their team or for the intermediate to advanced dbt developer who wants to become more efficient and learn a thing or two from software workflows. We'll talk about how automation can minimize the risk of code changes, shorten your time to merge, and improve quality and standardization throughout your projects.

Bring your laptop with dbt 1.0+ installed to follow along or just come and watch the walk-through! You'll start this workshop with a simple dbt project and end with local pre-commit automation on your computer and a framework for further improvement and iteration.

Check the step-by-step workshop guide here: https://blog.montrealanalytics.com/automating-dbt-development-workflows-with-pre-commit-b6c7ca708f7

Coalesce 2023 is coming! Register for free at https://coalesce.getdbt.com/.

Within Wehkamp we required a uniform way to provide reliable and on time data to the business, while making this access compliant with GDPR. Unlocking all the data sources that we have scattered across the company and democratize the data access was of the utmost importance, allowing us to empower the business with more, better and faster data.

Focusing on open source technologies, we've built a data platform almost from the ground up that focuses on 3 levels of data curation - bronze, silver and gold - which follows the LakeHouse Architecture. The ingestion into bronze is where the PII fields are pseudonymized, making the use of the data within the delta lake compliant and, since there is no visible user data, it means everyone can use the entire delta lake for exploration and new use cases. Naturally, specific teams are allowed to see some user data that is necessary for their use cases. Besides the standard architecture, we've developed a library that allows us to ingest new data sources by adding a JSON config file with the characteristics. This combined with the ACID transactions that delta provides and the efficient Structured Stream provided through Auto Loader has allowed a small team to maintain 100+ streams with insignificant downtime.

Some other components of this platform are the following: - Alerting to Slack - Data quality checks - CI/CD - Stream processing with the delta engine

The feedback so far has been encouraging, as more and more teams across the company are starting to use the new platform and taking advantage of all its perks. It is still a long time until we get to turn off some of the components of the old data platform, but it has come a long way.

Connect with us: Website: https://databricks.com Facebook: https://www.facebook.com/databricksinc Twitter: https://twitter.com/databricks LinkedIn: https://www.linkedin.com/company/data... Instagram: https://www.instagram.com/databricksinc/