Hey, I'm Ridwan. I am an Open Source Contributor and an AI Researcher.
Currently, I’m researching about AI and LLMs at the University of Debrecen in Hungary. I came to it from backend engineering, which turns out to be most of the job, the problems worth solving are inference latency, throughput, and the systems that hold up around a model.
Tech stack
GitHub activity
ProfileHover or tap the cards, months, legend or bars to explore the year.
Open source
All contributions35 pull requests merged across 6 repos
Projects
All projectsReads Amazon phone reviews, tells happy customers from unhappy ones, and shows sellers what people complain about most. A TF-IDF and logistic regression model tested on phones it had never seen, reaching a macro F1 of 0.918 and catching 92.6% of complaints.
A self-hostable backend-as-a-service in the spirit of Supabase: create a project and get its own Postgres schema, a REST API over its tables, S3 file storage with presigned uploads, realtime change events over WebSockets and per-project end-user auth, all managed from a web dashboard with a table and SQL editor.
Paste an email and see the verdict along with every word that argued for it, each shaded by how hard it pushed and which way. A multinomial Naive Bayes model over TF-IDF features, tuned by grid search and cut down by chi-squared selection, reaching 98.5% accuracy and 99.37% spam precision on 66,424 training emails.
A ride-hailing platform modeled on Uber: riders request trips, a matching service scores nearby drivers from Redis GEO search and reserves one atomically with a Lua script, and both sides follow the ride live on a map. Four Spring Boot services talk over Kafka with retries and dead-letter topics, behind Spring Cloud Gateway and STOMP over WebSocket.
Recent Blogs
All postsA plain-English story of how I taught a computer to read Amazon phone reviews, tell happy customers from unhappy ones, and show sellers what people complain about most.
Rebuilding the story of how Cloudflare cut about 100TB of RAM from its load balancer, step by step in Rust: checking every formula against real numbers and measuring what happens when the ring changes.
A simple breakdown of the paper “Self-Evolving Coding Agents via Harness-Policy Optimization”: a coding model that rewrites its own harness, trains inside it, and gains more than either fix alone.
Contact me
Let’s hop on a quick call and see if we’ve got the right chemistry.
Want to bounce ideas?
Let's jump on a quick call and explore what's possible.
Prefer email?
Mail is the fastest way to reach me, I reply within a day or two.
Budapest, Hungary
Ready to chat?
Book a free 20-minute call to talk through what you're building and see if it's a fit.
Book a Free Call“Turning your vision into digital reality is just one conversation away.”
Our hours on your clock · hover, drag or use ← →