AI & Data
AI Engineering roadmap
Build reliable LLM applications: prompting, tool calling, embeddings, RAG, agents, evaluation and security.
- Advanced
- about 12 hours
- 6 modules
- 13 steps
Before you start
What you will learn
1. ML Foundations
Supervised and unsupervised learning, training and validation, and the metrics used to judge models.
- Supervised & Unsupervised Learning — Machine Learning · Supervised & Unsupervised Learning
- Training, Validation & Cross Validation — Machine Learning · Training, Validation & Cross Validation
- Precision, Recall, F1 & ROC — Machine Learning · Precision, Recall, F1 & ROC
2. LLM Application Basics
How neural networks, transformers and LLMs work, tokens and context windows, and prompting for structured outputs.
- AI, Neural Networks & Transformers — AI Engineering · AI, Neural Networks & Transformers
- LLMs, Tokens & Context Windows — AI Engineering · LLMs, Tokens & Context Windows
- Prompt Engineering & Structured Outputs — AI Engineering · Prompt Engineering & Structured Outputs
3. AI-Assisted DevelopmentOptional
Work effectively with coding agents: give them context and instruction files, prompt for code, connect tools with MCP and review what they produce.
- Coding Agents — AI-Assisted Development · Coding Agents
- Context & Instruction Files — AI-Assisted Development · Context & Instruction Files
- Prompting for Code — AI-Assisted Development · Prompting for Code
- Tools & MCP — AI-Assisted Development · Tools & MCP
- Reviewing & Testing AI-Generated Code — AI-Assisted Development · Reviewing & Testing AI-Generated Code
4. RAG & Retrieval
Embeddings and vector search, and retrieval-augmented generation with chunking and reranking.
- Embeddings & Vector Search — AI Engineering · Embeddings & Vector Search
- RAG, Chunking & Reranking — AI Engineering · RAG, Chunking & Reranking
5. Tool Calling & Agents
Let models call functions and tools, and build agents with memory and MCP.
- Function & Tool Calling — AI Engineering · Function & Tool Calling
- Agents, Memory & MCP — AI Engineering · Agents, Memory & MCP
6. Production AI
Evaluate quality and hallucinations, defend against prompt injection, and manage latency, cost and routing.
- Evaluation & Hallucination — AI Engineering · Evaluation & Hallucination
- Prompt Injection & Guardrails — AI Engineering · Prompt Injection & Guardrails
- Latency, Cost, Routing & Observability — AI Engineering · Latency, Cost, Routing & Observability
Each step unlocks when you answer its questions well enough, and a module you finish counts in every roadmap that shares it.
Start this roadmap for free