AI Engineering

Engineer real-world LLM applications and multi-agent systems with LangChain,LangGraph, RAG, MCP, and vector databases

Apply now

To the programme

Duration

3 months (13 weeks)

Level

Intermediate

Requirements

30 learning hours per week

At least 1 year of coding experience with Python or JavaScript

Strong motivation to learn AI Engineering

Ability to participate during core learning hours (Mon–Fri, 09:00–16:30)

Good command of English (reading, writing, and comprehension)

What you’ll be able to do

LLM application development

Build and deploy real-world LLM applications using Python, LangChain, and LangGraph.

Prompt engineering, evaluation & APIs

Master prompt engineering and LLM evaluation while integrating multiple models — OpenAI, Anthropic Claude, Google Gemini, and Meta Llama — through OpenRouter.

RAG & vector databases

Use retrieval-augmented generation and vector databases like ChromaDB to connect models with external data.

AI agents

Design and deploy intelligent AI agents and multi-agent systems with LangGraph — using tool calling, MCP, and long-term memory to automate tasks and retain context.

Tools you'll learn

Python (or JavaScript)LangChainLangGraphRetrieval-augmented generation (RAG)LLM models from various providersClaude CodeMeta LlamaAnthropic ClaudeGoogle GeminiPrompt engineeringContext engineeringVector databases (e.g., ChromaDB)short-term and long-term memoryStreamlitAI agentsMCP

Curriculum

4 Modules · 15 Sprints

01Foundations of LLM Application Development4 sprints

Learn what it actually takes to build with large language models (LLM). This sprint gives you the technical and ethical foundations you need to start strong. You’ll explore prompt engineering, learn about LLM settings, and get hands-on with Python or JavaScript. You’ll also experiment with APIs like OpenAI, Anthropic, and others to learn how to craft and tweak effective prompts. By the end, you’ll have a good understanding of how LLMs work and how to use them with intention and impact for real-world applications.

1. Prompt engineering fundamentals

2. LLM settings and configuration

3. Python or JavaScript for LLM development

4. APIs: OpenAI, Anthropic, Google, Meta

02Building Applications with LangChain and RAG4 sprints

Here, you’ll discover how to build interactive tools. You’ll learn how to bring in external data using retrieval-augmented generation (RAG), work with vector databases like ChromaDB, and use LangChain to streamline LLM workflows. You’ll get practical experience building structured outputs, prototyping chatbots, and connecting LLMs to real-world data. By the end, you’ll have built your first full-stack LLM app using LangChain and Streamlit or Next.js.

1. Retrieval-augmented generation (RAG)

2. Vector databases: ChromaDB

3. LangChain for LLM workflows

4. Building chatbots with Streamlit or Next.js

03AI Agents4 sprints

Learn how to create AI agents that can perform complex, automated tasks. In this sprint, you’ll go from simple LLM chains to full agents that can make decisions, run code, call APIs, and carry out multi-step tasks. You’ll explore how memory works in agents so they can track context and perform more intelligently over time. With focused labs and a practical project, you’ll build agents that are useful, autonomous, and built for real-world use.

1. From LLM chains to full agents

2. Agent decision-making and API calls

3. Memory and context tracking

4. LangGraph for agent development

04Capstone ProjectCapstone · 3 sprints

Now it’s your turn to build something end-to-end. You’ll take on a self-directed capstone project where you design, prototype, and deliver a full LLM-powered application. Choose from real-world scenarios tied to different domains of AI engineering and make it yours.