Hello, I'm Dhyey Patel
I turn messy problems into clean, fast products — ensemble ML models, Django and FastAPI backends, and React interfaces that feel effortless. Computer Science undergrad, full-time builder.
I'm a Computer Science undergrad at Karnavati University's UnitedWorld Institute of Technology, based in Ahmedabad. I care about the unglamorous parts of software — clean data models, sensible APIs, systems that stay correct under load — and the polished parts people actually feel: speed, clarity, and interfaces that get out of the way.
Most of what I know came from shipping. Over the last two years I've built ERPs, SaaS platforms, ML pipelines and developer tools, usually end-to-end: schema to API to interface. If there's a hard problem with real users on the other side, I want in.
“Make it work, make it right, make it fast.” — Kent Beck
Understand the data, the constraints and the person on the other side before touching the keyboard.
Get one honest path working end-to-end early. Real feedback beats polished assumptions.
Measure, simplify, secure. Then sweat the details that make the product feel inevitable.
An internship running alongside the degree — real product work, real review cycles, no sandboxes.
Adrine · Core Product & Engineering
Karnavati University · UnitedWorld Institute of Technology
Nine specimens, catalogued from two years of shipping — data-heavy backends, AI in the loop, and interfaces that stay out of the way.
Specimens shown at natural scale — click one for the full field notes Tap a specimen for the full field notes
Preserved specimens — hackathon sprints, internal tools, and platforms that taught me something. All public, all built end-to-end.
Chosen for boring reliability, fast iteration, and the ability to go from idea to production without a committee.
Also in the toolbox — Postman · Linux · Alembic · PyMuPDF · pydub · Chart.js
A short list of things that actually happened — with dates, places and proof.
Won first place with Trackademic AI, an AI academic assistant that turns a student's deadlines into a prioritized study plan, tracks tasks and surfaces what to work on next. Built and demoed end-to-end inside the 36-hour window.
Built NexRail AI for Smart India Hackathon problem statement SIH26028: a dynamic ETA and decision-intelligence system for Indian Railways. It turns train running data and historical delays into arrival predictions, then surfaces the next operational decision for operators — from raw data to a working dashboard during the national hackathon.
Ten-plus end-to-end products across AI, SaaS, ERP and developer tooling — all designed, built and deployed solo.
I'm open to internships, new-grad roles, and ambitious side projects. If you're working on something interesting — or just want to trade notes — my inbox is open.