Learn directly from Cauchy

Nice learning journey! Guided by Maria and Basak, I went through most of the core aspects of LLM application development following LLMOps best practices. What impressed me most was seeing how seamlessly these capabilities are integrated within the Databricks ecosystem. Leveraging this opportunity, I was trying to build an AI agent for an internal use case to explore most of the essential components of agent development. Using the unified AI gateway, handling LLM endpoints on Databricks became much more straightforward. In addition, I equip the agent with tools exposed through MCP servers on Databricks, like Genie Space tools for text2sql and VSI tools for RAG workflows, and customized tools for the use case specifically. Agent memory management is realized using Lakebase. And for agent observability and evaluation, which are often cumbersome to implement from scratch, were handled in a much more manageable way through the MLflow framework. To make the demo agent more complete, I also implemented a simple state machine and agentic workflow orchestration, and finally deploy it as a serving endpoint on Databricks. Really impressed by all these capabilities within Databricks. Although some features are still evolving, the overall vision and ecosystem looks exciting. I also spent some time exploring deployment through Databricks Apps as it was mentioned as furture-proof way of deployment by Maria. Actually, Databricks provides a nice starter template with well-configured skills, and together with coding agents, AI agent development became much smoother and easier than I expected. Kudos to Maria and Basak for the guidance and support throughout the course! And looking forward to applying these learnings to build some interesting AI use cases to deliver real value in the production environments!

Very satisfied with the course overall. The instructors did an incredible job and were exceptionally available for questions and support throughout the program. I especially appreciated the practical focus on real production concerns around evaluations, monitoring, tracing, deployment, and infrastructure rather than just demos or notebooks.

comprehensive course with both in-depth theory and a lot of practical implementation. Most of the patterns and implementations are production-grade. Exploring Databricks features such as agent deployment, vector embeddings, Lakebase Progress SQL Server was very useful for understanding the agent lifecycle. Implementing CI/CD for Databricks deployments was also very interesting. Overall, the course is excellent. The quick support from the instructors and the weekly Q&A sessions helped me a lot.

Great course!

Great Course for someone trying to build agents in databricks env. I had a thorough walk through of concepts to be able to build out my metadata generator and agentic data contracts project.