Headline

Data Science Student | Building AI Solutions with Agentic AI, RAG & Computer Vision

About

Turning coffee, curiosity, and way too many open tabs into AI that actually does something useful. I’m a final-year Data Science student who’d rather build the thing than just read about it. So far that’s meant building AI agents that write documentation for you, generating exam papers with a bot that actually gets Bloom’s Taxonomy, and training a model to catch helmet violations from video. A few things I’ve broken (and eventually fixed) along the way: — DocBear: An agentic AI system that reads a codebase and rewrites the docs, so no dev has to — Qubit: a multi-agent tool that generates syllabus-aligned question papers, complete with answer keys and a human still checking things before they go out — A YOLOv8-powered system that catches helmet and seatbelt violations in real time — A full-stack “AI stock analyst” that reads live news and forms opinions faster than I do When I’m not training models, I’m probably at the gym, listening to music on loop, or planning the next trip. Always down to talk ML, AI, or good travel recommendations. Open to internships in ML/Data Science/AI Engineering — let’s build something.

Location

Bengaluru

AI stack

antigravity

claude code

claude

codex

cursor

deepseek

Education

Bachelor of Engineering, Computer Science and Engineering (Data Science), New Horizon College of Engineering

Pre-University, International Indian School, Dammam

Public links

github: https://github.com/ashlinalex1

leetcode: https://leetcode.com/ashlinalex