Skip to content

Repository files navigation

Databricks Mosaic AI Practice

A hands-on collection of Databricks notebooks exploring Generative AI and Mosaic AI capabilities, including Foundation Models, Embeddings, Retrieval-Augmented Generation (RAG), AI Assistants, Function Calling, AI Agents, Vector Search, and AI Gateway.

Overview

This repository contains practical notebooks created while learning and experimenting with Databricks Mosaic AI. The goal was to understand how modern LLM-powered applications are built using Databricks Foundation Models and related AI services.

The notebooks progress from basic model inference to more advanced concepts such as semantic search, retrieval pipelines, tool calling, agents, and AI governance.


Topics Covered

Foundation Models

  • Calling Databricks Foundation Models
  • Prompt-based interactions
  • Model inference using OpenAI-compatible APIs

Embeddings

  • Generating text embeddings
  • Understanding vector representations
  • Semantic similarity concepts

Retrieval-Augmented Generation (RAG)

  • Basic RAG workflow
  • Context-based question answering
  • Embedding generation and retrieval
  • Similarity search using cosine similarity

AI Assistants

  • System prompts
  • Custom AI assistants
  • Multi-turn conversations
  • Context retention across interactions

Function Calling

  • Tool definitions
  • External function execution
  • Structured outputs from LLMs

AI Agents

  • Multi-tool workflows
  • Dynamic decision-making using LLMs
  • Agent-style task execution

Vector Search

  • Semantic document retrieval
  • Query embeddings
  • Similarity matching
  • Context extraction

AI Gateway

  • Request rate limiting
  • Request logging
  • Governance and monitoring concepts

Repository Structure

Notebook Description
01_Foundation_Model_API Foundation Model inference using Databricks serving endpoints
02_Embeddings Generating embeddings with Databricks embedding models
03_RAG_Basics Introduction to Retrieval-Augmented Generation
04_AI_assistant Building a custom AI assistant
05_Multi-turn AI assistant Multi-turn conversational AI
06_Ai_function calling Function calling implementation
07_Dynamic user input Dynamic prompt handling
08_Dynamic Chatbot Interactive chatbot workflow
09_RAG End-to-end RAG implementation
10_Ai agent Agent-based AI workflow
11_Vector search Semantic retrieval using vector search concepts
12_Ai gateway rate limiting AI Gateway rate limiting simulation
13_Ai gateway logging AI Gateway request logging simulation

Technologies Used

  • Databricks
  • Mosaic AI
  • Foundation Models API
  • OpenAI Compatible SDK
  • Python
  • Vector Embeddings
  • RAG
  • Vector Search
  • AI Gateway

Learning Outcomes

By completing these notebooks, I gained practical experience in:

  • Working with Databricks Foundation Models
  • Generating and using embeddings
  • Building Retrieval-Augmented Generation (RAG) pipelines
  • Creating AI assistants and chatbots
  • Implementing function calling workflows
  • Developing basic AI agents
  • Understanding vector search and semantic retrieval
  • Applying AI governance concepts such as rate limiting and logging

Note

This repository was created for learning, experimentation, and understanding Databricks Mosaic AI concepts through hands-on practice.

About

A collection of Databricks notebooks demonstrating Generative AI workflows using Foundation Models, Embeddings, Vector Search, Retrieval-Augmented Generation (RAG), and AI Gateway.

Topics

Resources

Stars

0 stars

Watchers

0 watching

Forks

Releases

Packages

Contributors

Languages