This course provides a comprehensive review of DevOps principles and their application to Databricks projects. It begins with an overview of core DevOps, DataOps, continuous integration (CI), continuous deployment (CD), and testing, and explores how these principles can be applied to data engineering pipelines. The course then focuses on continuous deployment within the CI/CD process, examining tools like the Databricks REST API, SDK, and CLI for project deployment. You will learn about Databricks Asset Bundles (DABs) and how they fit into the CI/CD process. You’ll dive into their key components, folder structure, and how they streamline deployment across various target environments in Databricks. You will also learn how to add variables, modify, validate, deploy, and execute Databricks Asset Bundles for multiple environments with different configurations using the Databricks CLI. Finally, the course introduces Visual Studio Code as an Interactive Development Environment (IDE) for building, testing, and deploying Databricks Asset Bundles locally, optimizing your development process. The course concludes with an introduction to automating deployment pipelines using GitHub Actions to enhance the CI/CD workflow with Databricks Asset Bundles. By the end of this course, you will be equipped to automate Databricks project deployments with Databricks Asset Bundles, improving efficiency through DevOps practices. Pre-requisites: Strong knowledge of the Databricks platform, including experience with Databricks Workspaces, Apache Spark, Delta Lake, the Medallion Architecture, Unity Catalog, Delta Live Tables, and Workflows. In particular, knowledge of leveraging Expectations with Lakeflow Declarative Pipelines. Labs : Yes Certification Path: Databricks Certified Data Engineer Professional

This session is repeated.In this session, you will learn how to integrate Lakeflow Declarative Pipelines with external systems in order to ingest and send data virtually anywhere. Lakeflow Declarative Pipelines is most often used in ingestion and ETL into the Lakehouse. New Lakeflow Declarative Pipelines capabilities like the Lakeflow Declarative Pipelines Sinks API and added support for Python Data Source and ForEachBatch have opened up Lakeflow Declarative Pipelines to support almost any integration. This includes popular Apache Spark™ integrations like JDBC, Kafka, External and managed Delta tables, Azure CosmosDB, MongoDB and more.

What if you could run Spark jobs without worrying about clusters, versions and upgrades? Did you know Spark has this functionality built-in today? Join us to take a look at this functionality — Spark Connect. Join us to dig into how Spark Connect works — abstracting away Spark clusters away in favor of the DataFrame API and unresolved logical plans. You will learn some of the cool things Spark Connect unlocks, including: Moving you from thinking about clusters to just thinking about jobs Making Spark code more portable and platform agnostic Enabling support for languages such as Go

Ingesting data from SaaS systems sounds straightforward—until you hit API limits, miss SLAs, or accidentally ingest PII. Sound familiar? In this talk, we’ll share how Databricks evolved from scrappy ingestion scripts to a unified, secure, and scalable ingestion platform. Along the way, we’ll highlight the hard lessons, the surprising pitfalls, and the tools that helped us level up. Whether you’re just starting to wrangle third-party data or looking to scale while handling governance and compliance, this session will help you think beyond pipelines and toward platform thinking.

At Haleon, we've leveraged Databricks APIs and serverless compute to develop customer-facing applications for our business. This innovative solution enables us to efficiently deliver SAP invoice and order management data through front-end applications developed and served via our API Gateway. The Databricks lakehouse architecture has been instrumental in eliminating the friction associated with directly accessing SAP data from operational systems, while enhancing our performance capabilities. Our system acheived response times of less than 3 seconds from API call, with ongoing efforts to optimise this performance. This architecture not only streamlines our data and application ecosystem but also paves the way for integrating GenAI capabilities with robust governance measures for our future infrastructure. The implementation of this solution has yielded significant benefits, including a 15% reduction in customer service costs and a 28% increase in productivity for our customer support team.

Apache Spark™ has introduced Arrow-optimized APIs such as Pandas UDFs and the Pandas Functions API, providing high performance for Python workloads. Yet, many users continue to rely on regular Python UDFs due to their simple interface, especially when advanced Python expertise is not readily available. This talk introduces a powerful new feature in Apache Spark that brings Arrow optimization to regular Python UDFs. With this enhancement, users can leverage performance gains without modifying their existing UDFs — simply by enabling a configuration setting or toggling a UDF-level parameter. Additionally, we will dive into practical tips and features for using Arrow-optimized Python UDFs effectively, exploring their strengths and limitations. Whether you’re a Spark beginner or an experienced user, this session will allow you to achieve the best of both simplicity and performance in your workflows with regular Python UDFs.

Each LLM has unique strengths and weaknesses, and there is no one-size-fits-all solution. Companies strive to balance cost reduction with maximizing the value of their use cases by considering various factors such as latency, multi-modality, API costs, user need, and prompt complexity. Model routing helps in optimizing performance and cost along with enhanced scalability and user satisfaction. Overview of cost-effective models training using AI gateway logs, user feedback, prompt, and model features to design an intelligent model-routing AI agent. Covers different strategies for model routing, deployment in Mosaic AI, re-training, and evaluation through A/B testing and end-to-end Databricks workflows. Additionally, it will delve into the details of training data collection, feature engineering, prompt formatting, custom loss functions, architectural modifications, addressing cold-start problems, query embedding generation and clustering through VectorDB, and RL policy-based exploration.

Don’t you just hate telling your customers “No”? “No, I can’t get you the data that quickly”, or “No that logic isn’t possible to implement” really aren’t fun to say. But what if you had a tool that would allow you to implement those use cases? What if it was in a technology you were already familiar with — say, Spark Structured Streaming? There is a brand new arbitrary stateful operations API called TransformWithState, and after attending this deep dive you won’t have to say “No” anymore. During this presentation we’ll go through some real-world use cases and build them step-by-step. Everything from state variables, process vs. event time, watermarks, timers, state TTL, and even how you can initialize state with the checkpoint of another stream. Unlock your use cases with the power of Structured Streaming’s TransformWithState!

As data ecosystems grow more complex, organizations often struggle with siloed platforms and fragmented governance. In this session, we’ll explore how our team made Databricks the central hub for cross-platform interoperability, enabling seamless Snowflake integration through Unity Catalog and the Iceberg REST API. We’ll cover: Why interoperability matters and the business drivers behind our approach How Unity Catalog and Uniform simplify interoperability, allowing Databricks to expose an Iceberg REST API for external consumption Technical deep dive into data sharing, query performance, and access control across Databricks and Snowflake Lessons learned and best practices for building a multi-engine architecture while maintaining governance and efficiency By leveraging Uniform, Delta, and Iceberg, we created a flexible, vendor-agnostic architecture that bridges Databricks and Snowflake without compromising performance or security.

Spark Connect, first included for SQL/DataFrame API in Apache Spark 3.4 and recently extended to MLlib in 4.0, introduced a new way to run Spark applications over a gRPC protocol. This has many benefits, including easier adoption for non-JVM clients, version independence from applications and increased stability and security of the associated Spark clusters. The recent Spark Connect extension for ML also included a plugin interface to configure enhanced server-side implementations of the MLlib algorithms when launching the server. In this talk, we shall demonstrate how this new interface, together with Spark SQL’s existing plugin interface, can be used with NVIDIA GPU-accelerated plugins for ML and SQL to enable no-code change, end-to-end GPU acceleration of Spark ETL and ML applications over Spark Connect, with optimal performance up to 9x at 80% cost reduction compared to CPU baselines.

In this course, you'll learn concepts and perform labs that showcase workflows using Unity Catalog - Databricks' unified and open governance solution for data and AI. We'll start off with a brief introduction to Unity Catalog, discuss fundamental data governance concepts, and then dive into a variety of topics including using Unity Catalog for data access control, managing external storage and tables, data segregation, and more. Pre-requisites: Beginner familiarity with the Databricks Data Intelligence Platform (selecting clusters, navigating the Workspace, executing notebooks), cloud computing concepts (virtual machines, object storage, etc.), production experience working with data warehouses and data lakes, intermediate experience with basic SQL concepts (select, filter, groupby, join, etc), beginner programming experience with Python (syntax, conditions, loops, functions), beginner programming experience with the Spark DataFrame API (Configure DataFrameReader and DataFrameWriter to read and write data, Express query transformations using DataFrame methods and Column expressions, etc.) Labs: Yes Certification Path: Databricks Certified Data Engineer Associate

In this course, you’ll learn how to orchestrate data pipelines with Lakeflow Jobs (previously Databricks Workflows) and schedule dashboard updates to keep analytics up-to-date. We’ll cover topics like getting started with Lakeflow Jobs, how to use Databricks SQL for on-demand queries, and how to configure and schedule dashboards and alerts to reflect updates to production data pipelines. Pre-requisites: Beginner familiarity with the Databricks Data Intelligence Platform (selecting clusters, navigating the Workspace, executing notebooks), cloud computing concepts (virtual machines, object storage, etc.), production experience working with data warehouses and data lakes, intermediate experience with basic SQL concepts (select, filter, groupby, join, etc), beginner programming experience with Python (syntax, conditions, loops, functions), beginner programming experience with the Spark DataFrame API (Configure DataFrameReader and DataFrameWriter to read and write data, Express query transformations using DataFrame methods and Column expressions, etc.) Labs: No Certification Path: Databricks Certified Data Engineer Associate

Auto Loader is the definitive tool for ingesting data from cloud storage into your lakehouse. In this session, we’ll unveil new features and best practices that simplify every aspect of cloud storage ingestion. We’ll demo out-of-the-box observability for pipeline health and data quality, walk through improvements for schema management, introduce a series of new data formats and unveil recent strides in Auto Loader performance. Along the way, we’ll provide examples and best practices for optimizing cost and performance. Finally, we’ll introduce a preview of what’s coming next — including a REST API for pushing files directly to Delta, a UI for creating cloud storage pipelines and more. Join us to help shape the future of file ingestion on Databricks.

As data architectures evolve to meet the demands of real-time GenAI applications, organizations increasingly need systems that unify streaming and batch processing while maintaining compatibility with existing tools. The Ursa Engine offers a Kafka-API-compatible data streaming engine built on Lakehouse (Iceberg and Delta Lake). Designed to seamlessly integrate with data lakehouse architectures, Ursa extends your lakehouse capabilities by enabling streaming ingestion, transformation and processing — using a Kafka-compatible interface. In this session, we will explore how Ursa Engine augments your existing lakehouses with Kafka-compatible capabilities. Attendees will gain insights into Ursa Engine architecture and real-world use cases of Ursa Engine. Whether you're modernizing legacy systems or building cutting-edge AI-driven applications, discover how Ursa can help you unlock the full potential of your data.

In this course, you’ll learn how to define and schedule data pipelines that incrementally ingest and process data through multiple tables on the Data Intelligence Platform, using Lakeflow Declarative Pipelines in Spark SQL and Python. We’ll cover topics like how to get started with Lakeflow Declarative Pipelines, how Lakeflow Declarative Pipelines tracks data dependencies in data pipelines, how to configure and run data pipelines using the Lakeflow Declarative Pipelines. UI, how to use Python or Spark SQL to define data pipelines that ingest and process data through multiple tables on the Data Intelligence Platform, using Auto Loader and Lakeflow Declarative Pipelines, how to use APPLY CHANGES INTO syntax to process Change Data Capture feeds, and how to review event logs and data artifacts created by pipelines and troubleshoot syntax.By streamlining and automating reliable data ingestion and transformation workflows, this course equips you with the foundational data engineering skills needed to help kickstart AI use cases. Whether you're preparing high-quality training data or enabling real-time AI-driven insights, this course is a key step in advancing your AI journey.Pre-requisites: Beginner familiarity with the Databricks Data Intelligence Platform (selecting clusters, navigating the Workspace, executing notebooks), cloud computing concepts (virtual machines, object storage, etc.), production experience working with data warehouses and data lakes, intermediate experience with basic SQL concepts (select, filter, groupby, join, etc), beginner programming experience with Python (syntax, conditions, loops, functions), beginner programming experience with the Spark DataFrame API (Configure DataFrameReader and DataFrameWriter to read and write data, Express query transformations using DataFrame methods and Column expressions, etc.)Labs: NoCertification Path: Databricks Certified Data Engineer Associate

In this course, you’ll learn how to have efficient data ingestion with Lakeflow Connect and manage that data. Topics include ingestion with built-in connectors for SaaS applications, databases and file sources, as well as ingestion from cloud object storage, and batch and streaming ingestion. We'll cover the new connector components, setting up the pipeline, validating the source and mapping to the destination for each type of connector. We'll also cover how to ingest data with Batch to Streaming ingestion into Delta tables, using the UI with Auto Loader, automating ETL with Lakeflow Declarative Pipelines or using the API.This will prepare you to deliver the high-quality, timely data required for AI-driven applications by enabling scalable, reliable, and real-time data ingestion pipelines. Whether you're supporting ML model training or powering real-time AI insights, these ingestion workflows form a critical foundation for successful AI implementation.Pre-requisites: Beginner familiarity with the Databricks Data Intelligence Platform (selecting clusters, navigating the Workspace, executing notebooks), cloud computing concepts (virtual machines, object storage, etc.), production experience working with data warehouses and data lakes, intermediate experience with basic SQL concepts (select, filter, groupby, join, etc), beginner programming experience with Python (syntax, conditions, loops, functions), beginner programming experience with the Spark DataFrame API (Configure DataFrameReader and DataFrameWriter to read and write data, Express query transformations using DataFrame methods and Column expressions, etc.Labs: NoCertification Path: Databricks Certified Data Engineer Associate

The course intends to equip professional-level machine learning practitioners with knowledge and hands-on experience in utilizing Apache Spark™ for machine learning purposes, including model fine-tuning. Additionally, the course covers using the Pandas library for scalable machine learning tasks. The initial section of the course focuses on comprehending the fundamentals of Apache Spark™ along with its machine learning capabilities. Subsequently, the second section delves into fine-tuning models using the hyperopt library. The final segment involves learning the implementation of the Pandas API within Apache Spark™, encompassing guidance on Pandas UDFs (User-Defined Functions) and the Functions API for model inference. Pre-requisites: Familiarity with Databricks workspace and notebooks; knowledge of machine learning model development and deployment with MLflow (e.g. basic understanding of DS/ML concepts, common model metrics and python libraries as well as a basic understanding of scaling workloads with Spark) Labs: Yes Certification Path: Databricks Certified Machine Learning Professional

LLMs are magical—until they aren’t. Extracting adverse media entities might sound straightforward, but throw in hallucinations, inconsistent outputs, and skyrocketing API costs, and suddenly, that sleek prototype turns into a production nightmare.

Our adverse media pipeline monitors over 1 million articles a day, sifting through vast amounts of news to identify reports of crimes linked to financial bad actors, money laundering, and other risks. Thanks to GenAI and LLMs, we can tackle this problem in new ways—but deploying these models at scale comes with its own set of challenges: ensuring accuracy, controlling costs, and staying compliant in highly regulated industries.

In this talk, we’ll take you inside our journey to production, exploring the real-world challenges we faced through the lens of key personas: Cautious Claire, the compliance officer who doesn’t trust black-box AI; Magic Mike, the sales lead who thinks LLMs can do anything; Just-Fine-Tune Jenny, the PM convinced fine-tuning will solve everything; Reinventing Ryan, the engineer reinventing the wheel; and Paranoid Pete, the security lead fearing data leaks.

Expect practical insights, cautionary tales, and real-world lessons on making LLMs reliable, scalable, and production-ready. If you've ever wondered why your pipeline works perfectly in a Jupyter notebook but falls apart in production, this talk is for you.

The use of multiple Large Language Models (LLMs) working together perform complex tasks, known as multi-agent systems, has gained significant traction. While orchestration frameworks like LangGraph and Semantic Kernel can streamline orchestration and coordination among agents, developing large-scale, production-grade systems can bring a host of data challenges. Issues such as supporting multi-tenancy, preserving transactional integrity and state, and managing reliable asynchronous function calls while scaling efficiently can be difficult to navigate.

Leveraging insights from practical experiences in the Azure Cosmos DB engineering team, this talk will guide you through key considerations and best practices for storing, managing, and leveraging data in multi-agent applications at any scale. You’ll learn how to understand core multi-agent concepts and architectures, manage statefulness and conversation histories, personalize agents through retrieval-augmented generation (RAG), and effectively integrate APIs and function calls.

Aimed at developers, architects, and data scientists at all skill levels, this session will show you how to take your multi-agent systems from the lab to full-scale production deployments, ready to solve real-world problems. We’ll also walk through code implementations that can be quickly and easily put into practice, all in Python.

It started out as a fun weekend project, but as we know: it is never really that simple. In this talk, attendees can expect to learn the abbreviated version of everything we learned not to do when it comes to building and controlling remotely operated vehicles. We'll dive into the software and hardware side of things, talk about API design, why garbage collection is crucial and how we made all parts work together, reliably.