Gen AI Engineering with Databricks

This course is designed for data scientists, machine learning engineers, and data practitioners who want to build generative AI applications using Databricks. It provides hands-on experience with the latest frameworks and Databricks capabilities to develop, evaluate, and deploy generative AI solutions.

Participants will gain expertise in retrieval-augmented generation (RAG), multi-stage reasoning LLM chains, AI application governance, and model deployment. Through practical exercises and real-world use cases, learners will develop end-to-end generative AI applications while ensuring compliance and performance monitoring.

By the end of this course, learners will be able to:

• Implement retrieval-augmented generation (RAG) solutions using Databricks.
• Build and evaluate generative AI applications with multi-stage reasoning LLM chains.
• Apply governance and evaluation techniques to generative AI models.
• Deploy and monitor generative AI applications using Model Serving and Lakehouse Monitoring.
• Design autonomous AI agents with Databricks-based cognitive architectures.

Participants should have:

• Familiarity with natural language processing concepts.
• Understanding of prompt engineering and best practices.
• Experience with the Databricks Data Intelligence Platform.
• Knowledge of RAG concepts, including data preparation, embedding, vectors, and vector databases.
• Experience in building LLM applications using multi-stage reasoning LLM chains and agents.
• Familiarity with Databricks tools for AI evaluation and governance.

Target Audience

This course is intended for:

• Data scientists and machine learning engineers developing AI-driven applications.
• AI practitioners looking to enhance their skills in generative AI with Databricks.
• Organisations seeking to deploy and govern large-scale AI applications effectively.

Generative AI solution development

• Learn about RAG architecture and its role in generative AI
• Prepare data for RAG solutions to enhance contextual AI applications
• Explore embeddings, vector databases, and Mosaic AI Vector Search for vector search capabilities
• Assemble and evaluate a RAG application by integrating contextual information and assessing performance

Generative AI application development

• Understand how to decompose AI tasks and select the right models for business use cases
• Build multi-stage reasoning chains using LangChain and Hugging Face transformers
• Design and develop autonomous AI agents using generative models in Databricks

Generative AI application evaluation and governance

• Explore the importance of evaluating generative AI applications and ensuring governance and security
• Secure and govern generative AI applications using Databricks compliance tools
• Apply generative AI evaluation techniques to assess performance and effectiveness
• Conduct end-to-end application evaluations for cost, accuracy, and efficiency

Generative AI application deployment and monitoring

• Deploy generative AI models following industry best practices
• Implement batch processing workflows for AI model deployment
• Set up real-time deployment strategies for AI-powered applications
• Monitor AI systems to ensure performance and reliability using Databricks monitoring tools
• Apply LLMOps best practices to operationalise large language models

Exams and Assessments

This course does not include formal assessments.

Hands-On Learning

This course includes:

• Practical exercises with Databricks for generative AI model development.
• Hands-on labs covering RAG architecture, multi-stage reasoning, and agent design.
• Real-world AI governance and evaluation case studies.
• Guided model deployment and monitoring activities using Databricks tools.