← All roadmaps
Intermediate~10 weeks7 steps

AI Engineer

Build products on top of large language models: how LLMs work, calling model APIs, prompting, RAG, tool use, and evals. Best if you're already comfortable with Python or JavaScript.

  1. LLM fundamentals

    What a language model is: tokens, training, inference, and why they hallucinate.

  2. Working with model APIs

    Call Claude and other LLM APIs: messages, streaming, and multimodal inputs.

  3. Prompt engineering

    Get reliable outputs with roles, examples, structure, and chain-of-thought.

  4. Embeddings & RAG

    Ground models in your own data with embeddings, retrieval, and reranking.

  5. Tool use & agents

    Let models call tools and take multi-step actions safely.

  6. Evals & observability

    Measure quality systematically instead of eyeballing outputs.

  7. Production concerns

    Handle cost, latency, caching, safety, and failure modes at scale.