Materials Science & Engineering graduate building data products, applied ML systems, and engineering software.
I turn messy real-world data into reproducible, tested products — from ingestion and databases to APIs, models, and interfaces.
Currently building Touchline Intelligence, an end-to-end football analytics and decision-support platform built on open match data.
A full-stack football analytics platform that turns raw StatsBomb Open Data into a validated PostgreSQL database, a FastAPI backend, an analyst-facing Next.js interface, and a reproducible shot-quality modeling pipeline.
The project is built around a simple rule: data first, evidence second, model third.
Current work includes leak-free model cohorts, geometry and context features, time-aware evaluation, calibration, error analysis, and the progression from interpretable baselines toward gradient boosting and a PyTorch MLP.
Python · FastAPI · PostgreSQL · scikit-learn · Next.js · TypeScript · Docker
| Project | What it does |
|---|---|
| Touchline Intelligence | End-to-end football data platform with reproducible ingestion, quality audits, APIs, analytics, and applied ML. |
| MaterialScope | Python workbench for reproducible DSC, TGA, DTA, FTIR, Raman, and XRD characterization workflows. |
| PoroScope | Tested Python toolkit for calibrated porosity analysis in microscopy and SEM images. |
- Evidence before claims — measured outputs, explicit limitations, and reproducible records.
- Systems over demos — databases, APIs, interfaces, deployment, and validation belong together.
- Tests around real contracts — not just coverage numbers, but checks that protect intended behavior.
- AI-assisted, engineer-reviewed — I use modern coding agents aggressively, but decisions and verification stay human-owned.
Data & ML: Python, SQL, pandas, scikit-learn, PyTorch (in progress), Plotly
Backend & storage: FastAPI, PostgreSQL, REST APIs
Frontend: Next.js, React, TypeScript, Dash
Engineering: Docker, GitHub Actions, pytest, Ruff, mypy, Git
Domain background: materials characterization, microscopy, manufacturing, quality, and process engineering
My degree is in Materials Science & Engineering. That background shaped how I build software: measurement matters, provenance matters, assumptions must be visible, and a result is not trustworthy just because the code ran.
I'm currently looking for junior opportunities where engineering reasoning meets data, machine learning, Python/backend development, AI automation, or technical product work.
📍 İzmir, Türkiye