●3 reads●Creative Commons Attribution-NonCommercial-ShareAlike (CC BY-NC-SA)
RAG-Based AI Assistant Template: Complete Your Module 1 Project (AAIDC-Week4)
- AAIDC
- Agentic AI
- AI Assistant
- ChromaDB
- Document Retrieval
- LangChain
- Module 1 Project
- Project Template
- Project Tutorial
- RAG Template
- Vector Database
Table of contents
⬅️ Previous - Lesson 4 - RAG Implementation
➡️ Next - Project 1 Submission Guidelines
TL;DR
This is your first major project in the certification program. You’ll be building a Retrieval-Augmented Generation (RAG) assistant — an AI that can answer questions about your own documents.
To make the starting point less intimidating, we’ve prepared a project template, i.e., GitHub repo with scaffolded code, and a video walkthrough that shows you exactly how to complete it. Your job is to fill in a handful of missing functions that bring the system to life.
By the end, you’ll have a working Q&A assistant — think ChatGPT, but grounded in your knowledge base.
Why We’re Giving You a Template
Starting a big project from scratch can feel overwhelming. To help you past that “where do I begin?” moment, we’ve provided a ready-to-use template.
The template isn’t a complete project — instead, it gives you:
- A clear project structure with
app.py,vectordb.py, and adata/folder - Vector database integration (ChromaDB by default)
- Multi-LLM support (OpenAI, Groq, or Google)
- Sentence Transformers for embeddings
All the heavy lifting is in place. What’s missing are the core functions you’ll implement to make the assistant work.
What You’ll Build
Your task is to complete the system by adding the logic for:
- Loading your own documents
- Splitting them into chunks for search
- Creating embeddings and storing them in the vector database
- Searching for relevant chunks when a user asks a question
- Designing a prompt template that combines context with the user’s query
- Connecting it all into a query pipeline that produces an answer
In other words: the template gives you the skeleton, and you’ll add the muscles and movement.
🎥 Video Walkthrough: Implementing Your RAG Assistant
In this video, we guide you through the project template and show you exactly how to bring it to life. You’ll see how to:
- Set up your environment and API keys
- Add your own documents to the
data/directory - Implement the missing functions in
app.pyandvectordb.py - Choose chunking strategies and embedding models
- Run queries and test your assistant end to end
What You’ll Build
By completing the template, you’ll have an AI assistant that can:
- Load and process your own documents
- Break them into chunks for semantic search
- Store and retrieve them from a vector database
- Use an LLM to generate contextual answers
What’s Already Done for You
The template handles:
- Multi-LLM support (OpenAI, Groq, Google)
- ChromaDB integration, ready out of the box
- Sentence Transformers embeddings
- Clean structure and environment setup
All the infrastructure is ready. You just add the core logic.
Success Criteria
You’ve completed the project when:
- You can add your own documents to the
data/folder - The system loads, chunks, embeds, and stores them
- Queries return relevant results
- The assistant generates meaningful, contextual answers
- You can switch between LLM providers via environment variables
Ready to Start?
Clone the repo, watch the walkthrough, and implement one function at a time. Test as you go — and remember, the template is here to help you overcome the blank-page problem.
This is your first step toward building real agentic systems. Have fun with it! 🚀
⬅️ Previous - Lesson 4 - RAG Implementation
➡️ Next - Project 1 Submission Guidelines