Complete Generative AI Bootcamp 2026: LangChain, Agents, RAG

Discover the ultimate 2026 generative ai roadmap, covering ai overview, branches, generative ai models (llms, vlms, slms), langchain, olama, mcp, a2a, rag, and fine-tuning.

Explore Hugging Face to discover model categories for text generation, image generation, and text-to-video, access models via api keys and inference providers, and browse datasets and spaces.

Explore the LangChain framework by building LLM workflows with chaining, tools, and LCEL, guided by scenario-based learning, and prepare to set up a Python-driven AI assistant using API keys.

Define the problem and framework, then follow the golden rule to configure an LLM with LangChain. Set temperature, craft prompts, select a pattern, build pipeline, and deploy as an API.

Create and activate a conda Python environment, configure OpenAI and Hugging Face keys in a .env file, and install LangChain dependencies for a complete generative AI setup.

Define the problem, choose LangChain as the framework, select models, configure the LLM, write prompts, assemble the chain, tools, and memory to create your AI agent.

Select and configure OpenAI paid and HuggingFace free language models using LangChain. Load the API key from .env, set temperature and max tokens, and save as llm_open_ai and llm_hugging_face.

Compare OpenAI GPT-5 and Hugging Face's DeepSeq by initiating chats, invoking models, and evaluating initial responses to test large language models in a hands-on demo.

Develop the prompt template in LangChain's golden rule step 5 to endow the AI with a personality. Create and load prompt templates for OpenAI and HuggingFace models using LangChain.

Build your first simple LLM chain using LangChain patterns, loading OpenAI and HuggingFace models, defining a prompt template, and invoking the chain for Q&A.

In this step 6 lecture, extend a simple chain with LangChain tools and an LLM, enabling tool calling for rating, food information, and opening hours, and introducing a React agent.

Learn to build a react agent with thinking, action, and observing capabilities using LangChain, integrating multiple tools for multi-step AI reasoning.

Build LangChain pipelines with LCEL by initializing LLMs, defining prompts, and composing a pipeline with the pipe operator, including Hindi prompts and parsing to Hindi outputs.

Add memory to an autonomous agent using LangChain, with checkpoint memory and an in memory server to remember conversations across per-user threads tied to a thread ID.

Build a complete food agent using LangChain memory tools and chaining to answer food queries, recipes, nutrition, and restaurant information with memory across sessions.

Learn why running LLMs locally offers offline, secure AI with OLAMA, and how to pull and run models and build conversations with LangChain, without API keys.

Install ollama on mac, windows, or linux and access a library of models. Download and run a local model (llama 3.1) offline via pull, list, run, ps, stop, and rm.

Learn to run local AI apps with LangChain and Ollama by configuring Ollama models, writing prompts, and building tools, agents, and memory for offline, fully local conversations.

Explore the Model Context Protocol (MCP) as a universal connector that links diverse tools, databases, and browsers via Docker, enabling intelligent LLM-driven tool selection.

Install Docker Desktop, set up the MCP toolkit, and run MCP tools like YouTube transcribe and Playwright inside containers to power a client-server AI agent architecture with cursor.

Discover the Google agent development kit, a production-ready tool for autonomous agents. Learn to build multiple agents, use a pre-configured chat bot with memory, and run locally or via API.

Build a multilingual menu assistant in three languages using the Google ADK, LangChain tools, and OLAMATI local models, plus cloud kitchen agents with order tools.