Back to course
Download flier
GrowTechSkill Course Flier · 2026

LLMs

Hands-on LLM engineering - from prompting to agents and MCP servers.

8 weeks Saturdays & Sundays 13:00–16:00 (UTC+1) Live online
Free demo & orientation
Thu, 19 November 2026 · 13:00–16:00
Course runs
22 November – 11 January 2027
+3 days SatSun2 months

What you'll learn

  • Understand how LLMs work - transformers, tokens and context windows
  • Master prompting techniques that get reliable, structured output
  • Build RAG pipelines with vector stores (pgvector, Qdrant, ChromaDB)
  • Create conversational chatbots with memory and context
  • Build AI agents with LangChain and LangGraph
  • Deploy MCP servers and connect agents to external tools
  • Evaluate LLM applications for quality, cost and safety
  • Ship a portfolio-ready LLM project

8 weeks to job-ready LLMs

Week 1 · LLM foundations What are LLMs - transformers, tokens, context windows, Setting up: API keys, environment, first API call, Prompting fundamentals and best practices…
Week 2 · Prompting and text generation Advanced prompting: few-shot, chain-of-thought, Structured output: JSON mode and function calling, Building a CLI chatbot with the OpenAI / Anthropic API…
Week 3 · RAG - Retrieval-Augmented Generation What is RAG and why it matters, Document loading, chunking strategies and embedding, Vector stores: pgvector, ChromaDB, Qdrant…
Week 4 · Embeddings and vector stores deep dive Embedding models: OpenAI, open-source alternatives, Similarity search, filtering and hybrid strategies, pgvector with PostgreSQL in production…
Week 5 · LangChain fundamentals LangChain architecture: chains, prompts, output parsers, Sequential and parallel chains, Memory and conversation history…
Week 6 · LangGraph agents What are AI agents and when to use them, LangGraph: state machines for multi-step reasoning, Tool-calling agents with persistent memory…
Week 7 · MCP servers and integration Model Context Protocol (MCP) - architecture and spec, Building an MCP server from scratch, Connecting agents to external tools via MCP…
Week 8 · Capstone - build your LLM app Design your LLM-powered application, Build it hands-on, session by session, Deployment, monitoring and production safety…

You'll use

Python 3.12+VS CodeOpenAI APIAnthropic APILangChainLangGraphpgvectorQdrantDockerGit and GitHub