Why a ramen stall and not a portfolio template
What an interactive front door buys you, and where it just gets in the way.
Everything here was built by the cook. Ask the guide about any of it.
The big builds first, then the small sharp tools, then the ventures and the research.
Machine-learning study of used spare-part pricing. In review and defence.
Student assistant: agent internals and developer-experience work.
A deliberate ground-up rebuild of the maths and stats under all of it.
the builds I most like explaining
A chess AI trained only on my own Chess.com history. It encodes each move as a compact tuple and predicts my next move, imitating how I actually play rather than how an engine would. Source code ↗ · Live demo ↗
Real-time messaging: private and group rooms, read receipts, friend requests, GDPR data controls. Django Channels over WebSockets with Redis pub/sub, deployed on Railway with PostgreSQL. Source code ↗
focused tools and experiments
A scheduled Airflow pipeline that extracts, cleans and validates weather data, engineers daily and monthly aggregations plus wind categories, and flags outliers by threshold before loading to SQLite. Source code ↗
An endless runner you control with your body. Pose estimation reads the webcam and turns lean, jump and duck movements into game input in real time. Source code ↗
A neural net predicting term-deposit sign-ups: 89.3% accuracy and 0.76 AUC over 41,188 records, with batch normalisation and dropout to handle heavy class imbalance. Source code ↗
A small local Mac tool that turns messy copied text into clean, structured, reusable data. Narrow scope, does one thing. Source code ↗
ventures and research
An AI Finnish speaking-practice platform, built through Boost Turku's Startup Journey. Generalist founder role: I built the marketing site and the app prototype. Live demo ↗
Machine-learning analysis of used automotive spare-part pricing: what drives price, and estimating market value from raw listing data. Final-year thesis, in review. Source code ↗
More detail, and the parts that aren’t on GitHub, coming with the full site.
Not the cook. The one who tells you what the cook has been up to.
Pull up a stool. The cook here is Ritesh, final year Data & AI Engineering at Turku University of Applied Sciences, and a student assistant at a research group called Core Cognitive Tech.
He likes the unglamorous half of machine learning. Not the demo: the part where a model or a data pipeline has to run in production without waking anyone at three in the morning.
Ten years of football, one recent first half marathon, English and Finnish, learning Swedish. Everything on the menu he built himself, so ask about any of it.
Short notes on what got built and what broke.
Working notes, technical breakdowns and the decisions behind the projects.
Open the complete log →What an interactive front door buys you, and where it just gets in the way.
Move encoding, a tiny vocabulary, and what "plays like me" actually measures.
The short version. Take it with you.
The full CV as a PDF arrives once the copy is final. Until then the links are the source of truth.
Total: one hire. — service not included
Draft copy, written by the assistant. Ritesh replaces it in his own voice.