Course Wiki

Prompt Engineering

Designing the instructions and context an LLM responds to: prompt structure, examples, and iteration against real outputs.

LLM Zoomcamp treats the prompt as the core programmable surface of an LLM application: a template that combines instructions, the retrieved context, and the user question. Module 1 builds the RAG prompt step by step — role and instructions, context section, question section — and shows how each part affects answer quality.

The course’s evaluation module closes the loop: prompt variants are A/B-compared on a golden question set instead of judged by vibes, which is the discipline that separates prompt engineering from guessing. Token limits and API costs shape how much context a prompt can carry.

Taught in

• Machine Learning Zoomcamp

  • Module 8: Neural Networks and Deep Learning
    • Installation of tensorflow

• LLM Zoomcamp

  • Module 1: Agentic RAG
    • Introduction
    • RAG
    • Building the Prompt
    • The LLM
    • RAG Helper
    • Data Ingestion
    • Wrap-up of Part 1
    • Agents
    • Quick RAG Revision (Optional)
    • Function Calling
    • The Agentic Loop
    • ToyAIKit
    • Other Frameworks
  • Module 2: Vector Search
    • RAG with Vector Search
  • Module 3: Orchestration
    • Context Engineering
    • AI Copilot
    • Retrieval Augmented Generation
    • AI Agents
    • Multi-Agent Systems
    • Best Practices
  • Module 4: Evaluation
    • Evaluation
    • Search Parameter Tuning
    • RAG and Agent Evaluation
    • Generating RAG Answers
    • LLM as a Judge
    • Agent Evaluation
    • Next Steps
  • Module 5: Monitoring
    • Monitoring
    • Assistant
    • Capturing Metrics
    • Storing Data in PostgreSQL
    • Querying Data
    • Streamlit Dashboard
    • User Feedback
    • Built-in Judge
    • Synthetic Data Generation
  • Module 7: End-to-End Project
    • End-to-End Project Example
    • Evaluating Retrieval
    • Evaluating RAG
    • Interface and Ingestion Pipeline
    • Monitoring and Containerization

• AI Dev Tools Zoomcamp

  • Module 4: DevOps and Observability for AI-Built Apps
    • DevOps and Observability for AI-Built Apps
  • Module 5: Coding Agent Building Blocks: Reusable Skills and Specialized Subagents
    • Module 5 — Coding Agent Capabilities: MCP, Skills, Plugins, and Custom Agents ## Related concepts

• RAG

• Function Calling

• LLM Evaluation

• OpenAI API

• RAG