KenteCodeGPT: Building a Production-Ready RAG System

Β· 1 min read

πŸš€ Project #1 for Our Generative AI Engineering Cohort

The best way to learn Generative AI is to build real systems.

For the first project in our Generative AI Engineering program, students will design and implement a Retrieval-Augmented Generation (RAG) application: the KenteCode Course Assistant.

The goal is to create an AI assistant capable of answering natural language questions about KenteCode courses, learning paths, fees, schedules, and policies while grounding every response in approved source documents.

The project will cover:

  • βœ… Document ingestion and processing
  • βœ… Text chunking and embeddings
  • βœ… Vector search with Pinecone
  • βœ… Retrieval-Augmented Generation (RAG)
  • βœ… Prompt engineering and hallucination control
  • βœ… FastAPI backend development
  • βœ… Streamlit chat interface
  • βœ… Source citations and answer grounding

Proposed Architecture

πŸ“„ Documents β†’ Chunking β†’ Embeddings β†’ Pinecone

❓ User Query β†’ Retrieval β†’ LLM β†’ Grounded Answer + Sources

Tech stack:

  • Python
  • OpenAI Embeddings
  • Pinecone
  • FastAPI
  • Streamlit
  • PyPDF and document parsers
  • Recursive text splitting

Students won't just learn about RAG systemsβ€”they will build one similar to the AI assistants powering many modern enterprise applications.

This is exactly the type of hands-on, portfolio-worthy experience that prepares engineers to develop production-ready AI applications.

Looking forward to seeing what the cohort builds!

Share this post