Skip to main content
Iroko Skills Academy
Menu

AI & Application Development

Building LLM Apps (RAG and Prompting in Code)

Build software powered by language models: calling an LLM API in code, designing reusable prompts, grounding answers in your own documents with Retrieval-Augmented Generation (RAG), giving the model tools via function calling, and evaluating quality. Ends with a RAG build capstone.

ADVANCED · en · ~360 min

Instructor: Self-paced · Reviewed by: Iroko Skills Academy team · Published by: Iroko Skills Academy team

Learning objectives

  • Build the core LLM-app loop: assemble prompt, call the API, use the response
  • Design reusable system/user prompt templates in code
  • Implement Retrieval-Augmented Generation to ground answers in your documents
  • Add tool/function calling with guardrails and evaluate app quality

Before you start

Prerequisites: None

Device/connectivity: Laptop or desktop computer with a modern web browser and keyboard, Internet access for lessons, tools, and assignment submission

Sign in to enroll

What you'll cover

  1. From Chat to App

    • From using an LLM to building with one
    • Prompts in code: system, user, and templates

  2. Grounding and Tools

    • How retrieval works: embeddings and similarity
    • Giving the model tools (function calling)
    • Grounding in code: RAG

  3. Quality and Capstone

    • Evaluating and improving an LLM app
    • Capstone: build a small RAG application

Assessment and completion criteria

Complete every published lesson to earn a completion badge. A self-paced certificate also requires passing every server-graded quiz. Practical assignments provide project feedback and may support a separate verified-proficiency record, but they do not delay the standard course certificate.

  • Optional reviewed project: Build a small RAG application · rubric v1 · pass ≥ 70%