Master Langchain v1 and Ollama - Chatbot, RAG and AI Agents

Descripción

2026 Upgrade: Course completely re-recorded with LangChain v1 and LangGraph v1. All projects, agents, tools, and RAG pipelines rebuilt from scratch. **Perfect for developers, AI engineers, and serious learners who want production-grade GenAI skills.** This course is a comprehensive, practical guide to integrating Langchain v1 (latest release) and Ollama to build, automate, and deploy production-ready AI applications. Updated with the newest technologies and frameworks, you'll learn to set up these cutting-edge tools, create advanced prompt templates, build autonomous AI agents, implement RAG (Retrieval-Augmented Generation) systems, and deploy real-world applications on AWS. Each section is designed to provide you with hands-on skills and real-world experience with the latest AI development practices. What You Will Learn 1. Ollama & Langchain Setup • Complete installation and configuration of Ollama and Langchain • Work with the latest models: GPT-OSS, Gemma3, Qwen3, DeepSeek R1, and LLAMA 3.2 • Master Ollama commands, custom model creation, and raw API integration • Configure local LLM environments for optimal performance 2. Advanced Prompt Engineering • Design effective AI, human, and system message prompts • Use ChatPromptTemplate and MessagesPlaceholder for dynamic conversations • Master the invoke method and structured prompt patterns • Implement best practices for prompt tuning and optimization 3. LCEL Chains for Workflow Automation • Build Sequential, Parallel, and Router Chains with Langchain Expression Language (LCEL) • Create custom chains using RunnableLambda and RunnablePassthrough • Implement chain decorators for simplified workflow automation • Design conditional logic and dynamic chain routing for complex applications 4. Structured Output Parsing • Parse LLM outputs using Pydantic, JSON, CSV, and custom parsers • Use with_structured_output method for type-safe responses • Handle date-time parsing and structured data extraction • Format data for downstream processing and integration 5. Chat Memory and Conversation Management • Implement chat history with BaseChatMessageHistory and InMemoryChatMessageHistory • Use MessagesPlaceholder for dynamic conversation flow • Build stateful conversational AI applications • Manage long-term chat sessions efficiently 6. Build Production-Ready Chatbots • Create interactive chatbot applications using Streamlit • Implement streaming responses like ChatGPT • Maintain persistent chat history and session state • Deploy user-friendly chat interfaces with real-time updates 7. Document Processing with Multiple Loaders • Process PDFs using PyMuPDF and create QA systems • Work with Microsoft Office files (PPTX, DOCX, Excel) • Use Microsoft's MarkItDown for universal document conversion • Implement IBM's Docling for advanced OCR and document processing • Extract tables, images, and figures from any document type 8. Vector Stores and RAG Implementation • Build Retrieval-Augmented Generation (RAG) systems with FAISS and Chroma • Create and manage vector embeddings using OllamaEmbeddings • Implement document chunking strategies with RecursiveTextSplitter • Optimize chunk sizes for better retrieval performance • Design RAG prompt templates for context-aware responses 9. Agentic RAG Systems • Build autonomous RAG agents that retrieve and reason • Create custom tool decorators for agent capabilities • Implement real-time streaming for agent responses • Integrate vector stores with intelligent agent workflows 10. Tool Calling and Function Execution • Set up built-in tools: Tavily Search, DuckDuckGo, PubMed, Wikipedia • Create custom tools and bind them to LLMs • Implement tool calling loops for multi-step reasoning • Pass tool results back to LLMs for informed responses 11. AI Agents with Langchain • Master the create_agent API for building intelligent agents • Build web search agents with DuckDuckGo integration • Implement agent state management and middleware • Create dynamic model selection for intelligent agent routing • Stream agent responses in real-time using values, updates, and messages 12. Text-to-SQL Agent (MySQL Integration) • Build natural language to SQL query systems • Create schema inspection, query generation, and validation tools • Implement automatic SQL error correction with LLMs • Execute complex database queries from natural language 13. Real-World AI Projects • Stock Market News Analysis: Scrape web data and generate comprehensive reports • LinkedIn Profile Scraper: Extract and parse profile data with LLMs • Resume Parser: Build AI-powered CV analysis and JSON extraction system • Health Supplements QA: Create domain-specific RAG question-answering systems 14. Production Deployment on AWS • Launch and configure AWS EC2 instances for LLM applications • Install Ollama and Langchain on cloud servers • Deploy Streamlit applications in production environments • Connect VS Code to remote servers for seamless development By the end of this course, you'll have the expertise to build, deploy, and manage production-grade AI-powered applications using Langchain and Ollama. You'll be able to create intelligent chatbots, RAG systems, autonomous agents, and document processors that are ready for real-world deployment. Start building the future of AI applications today.

Lo que aprenderás

• Install and integrate LangChain v1 and Ollama to run Qwen3, Gemma3, DeepSeek R1, GPT-OSS, LLAMA, and custom GGUF models locally.
• Build complete chatbots with memory, history, streaming responses, and a Streamlit UI.
• Use prompt templates, LCEL chains, chain routing, parallel chains, custom chains, and runnable pipelines to structure LLM workflows.
• Parse structured output using Pydantic, JSON, CSV parsers, and .with_structured_output() methods.
• Implement advanced retrieval systems including similarity search, MMR search, threshold search, and optimized chunking.
• Use tool calling and function calling with DuckDuckGo, Tavily, Wikipedia, PubMed, and custom tools.
• Build production-ready AI agents using LangChain v1 agent API, dynamic model selection, middleware, state management, and real-time streaming.
• Create Agentic RAG systems including autonomous retrieval, context citation, custom FAISS tools, and streamed agentic responses.
• Build a complete Text-to-SQL Agent for MySQL with schema extraction, SQL generation, validation, execution, and automated error correction.
• Build LinkedIn scraper, resume parser, and data extraction workflows using Selenium, BeautifulSoup, LLM parsing, and Streamlit apps.
• Deploy LangChain v1 + Ollama applications to AWS EC2, configure remote servers, and run production-level AI apps.

Requisitos

• Basic Python programming knowledge
• Familiarity with APIs and web requests
• Basic understanding of machine learning concepts
• Access to a computer with internet for installations and setups
• Curiosity to learn LLMs, AI agents, and RAG systems — everything else will be taught step-by-step.