Hello ! 
I’m excited to share my latest project: Aka Cleaner.
As a recent Computer Science graduate from Mulungushi University in Zambia, I wanted to build a tool that solves the “black box” problem in data cleaning. Most tools just fix data without explaining why.
Aka Cleaner is an Agentic Data Assistant that doesn’t just automate tasks—it reasons with the user.
Key Features:
-
AI Reasoning Engine: Explains the impact of every cleaning step (e.g., “Missing values in Target may bias your model”).
-
Mentor Mode: A dedicated educational toggle to help beginners learn the logic behind data prep.
-
Human-in-the-Loop: Every suggestion requires user approval, ensuring transparency.
-
Interactive EDA: Built using Plotly for instant visual health checks.
I built this using Streamlit because of its incredible ability to turn complex Python logic into a user-friendly UI. I’d love to get your feedback, especially on the Agentic reasoning flow!
Live App: https://aka-cleaner.streamlit.app/ GitHub Repo: GitHub - LOSTBAGWELL/agentic-data-cleaner: An intelligent, human-in-the-loop data cleaning engine that uses agentic reasoning to diagnose, validate, and transform messy datasets into model-ready features.
Built with 
in Zambia.
