Mission 8 of 12 · Updated November 2026

Build an AI Recruitment Assistant with RAG

Put an LLM inside your product. Call the OpenAI Responses API from Express, engineer prompts with instructions, examples and constraints, validate structured JSON with Zod, answer questions from a document collection with embeddings-based RAG, and trace every model call in Langfuse. You also start directing Codex CLI as your coding agent, with an AGENTS.md that sets the rules.

By the end of this mission: “I can build a reliable LLM feature with prompt and context engineering, structured output, basic RAG and tracing.”

The tech stack for this mission

OpenAI logoOpenAINode.js logoNode.jsExpress logoExpressZod logoZodLangfuse logoLangfuseCodex CLI logoCodex CLI

What you learn, checkpoint by checkpoint

  1. Checkpoint A

    Call the OpenAI Responses API and engineer reliable prompts

    • How LLM APIs work: inputs, outputs, tokens and context windows
    • Call the OpenAI Responses API from your Express backend
    • Prompt engineering: instructions, examples and constraints

  2. Checkpoint B

    Return validated structured output and answer from retrieved documents (RAG)

    • Get structured JSON from the model and validate it with Zod
    • Context engineering: select and supply only the information the model needs
    • Build basic RAG with the OpenAI Embeddings API over a small document collection

  3. Checkpoint C

    Trace calls in Langfuse, write repeatable test cases, and direct a coding agent

    • Trace model calls, latency, token usage and cost in Langfuse
    • Write at least five repeatable test cases for your LLM output
    • Direct Codex CLI as a coding agent, with an AGENTS.md for instructions, context and checks

AI companion mode · AI as Product

AI runs inside what you build

AI lives inside the product you build. Treat its output as untrusted - validate it, trace it, evaluate it, and design for uncertainty.

What proves you’re done

Working AI assistant + validated JSON output + basic RAG + five repeatable test cases + a Langfuse trace

Every mission ships something real you can show. If you can’t explain it, you can’t submit it.

Deliberately left out of this mission

Custom vector database infrastructure · Advanced RAG pipelines · Fine-tuning · Advanced prompt optimisation

Depth is set on purpose: you master what this mission needs, not everything at once. The advanced material lives in optional self-study after Mission 12.

Free · no card needed

Reading is not building.

This mission is part of the paid 12-week programme - but the way we teach is free to try. The free trial puts you inside Sigmo, the learning portal our students use daily, with the 6 Projects in 6 Days precourse: real projects, real code, published to your own GitHub.

Start the free trial

Mission 7

Test, Review & Deploy the Recruitment Platform

Mission 9

Build an AI Recruitment Agent with Tools & MCP

Ready for the real thing?

The AI-Native Software Development Programme - all 12 missions with a live cohort, daily Buildroom, and mentors watching your back.

Apply nowSee the whole curriculum