Building My First Q&A System With VS Code, Python, FastAPI & ChromaDB β
documenting the real journey of learning, building, breaking, and trying again.
Video Walkthrough
π Why I Started This Project
Iβve been sharing my journey into AI engineering through short video updates, but I wanted to also document the deeper learning behind it.
This project started because I wanted to learn:
-
how AI apps are structured
-
how retrieval works
-
how a pipeline goes from user β query β embeddings β vector store β response
-
how to glue together Python, FastAPI, and ChromaDB
-
how to build real things (not just follow tutorials)
I built a Q&A document system where you can upload text and ask questions, and the system retrieves answers using embeddings and a vector database.
This blog post breaks down everything I learned while building it.
π οΈ What I Built
I created a small but complete system using:
-
Python
-
FastAPI
-
OpenAI API
-
ChromaDB for embeddings
-
VS Code as my development environment
-
A simple front-end interface to ask questions
The flow is:
-
User uploads a document
-
The system creates embeddings
-
ChromaDB stores them
-
When the user asks a question, the system retrieves relevant chunks
-
OpenAI generates a final answer
This is the foundation of RAG (Retrieval-Augmented Generation) β something all AI engineers eventually learn.
π What I Learned (and What You Can Learn Too)
1. Pipelines Matter
This was one of the biggest mindset shifts.
AI apps are pipelines, not magic.
You need to think:
-
How does the input move?
-
What transformations happen?
-
What data is stored?
-
How does retrieval work?
-
How do you evaluate the output?
Once you grasp that, everything becomes much clearer.
2. Documentation Helps You Think
In my videos, I showed my README.md file because documenting your process:
-
helps you stay organized
-
helps others follow your thinking
-
builds your public learning journey
-
turns your project into content
-
makes you more confident explaining your work
3. Making Mistakes Is the Point
In the third video, I encouraged others to try, break things, fix things, and keep learning.
AI engineering is not about being perfect.
Itβs about iteration and curiosity.
π§© Project Structure
(You can also include this directly from your README)
Β project/
|ββ app/
|ββ routes/
|ββ models/
|ββ services/
|ββ chromadb/
|ββ fastapi_app.py
|ββ requirements.txt
|ββ README.md
This structure helped me understand how real AI apps are built.
π± Why I’m Sharing This Journey
Iβm building SheTechy, a brand and community for women who want to learn tech, build projects, and create their own opportunities.
Documenting your journey is powerful because:
-
It shows others the process
-
It builds your reputation
-
It attracts collaborators
-
It helps you build content at scale
-
It becomes your digital portfolio
These small videos and posts are the seeds of something much bigger.
π¬ Final Thoughts
This is just the beginning.
My next steps are:
-
expanding this AI app
-
exploring more RAG techniques
-
building more tools in public
-
continuing to share every step
If you want to follow, learn, or build along with me, stay connected with SheTechy.


