Local intelligence
A local-model enrichment pipeline: crawl public company sites and extract industry and market-positioning data with Qwen.
Available for the next challenge
Wrocław, PL / 51° N 17° EHÜSEYİN TUNAY ÇELİK ↗
Software engineer at the intersection of
AI, full-stack products & cloud systems.
From an idea to something that works.
Public company websites become the input.
~110 WEBSITES · SECURE SCRAPINGRead the project ↗01 Data: Public company websites become the input.
01 / Selected work
A selection of systems I’ve built.
The thinking behind the work, included.
A local-model enrichment pipeline: crawl public company sites and extract industry and market-positioning data with Qwen.
A unified operations panel for listings, orders, warehouse locations and integration health across Turkish marketplaces.
Frontend ownership of a real-estate CRM and AI marketing studio, from a typed API client to an editable visual-composition workflow.
PDF generation, AI tailoring and AWS storage brought together in a Flask application with pre-signed document access.
A deterministic campaign engine for a scan-based inventory system: time windows, explicit priorities and an auditable price history.
02 / Research in progress
THESIS · AZURE INCIDENT RESPONSEHow much of incident response can we automate—and where does the time actually go?
A repeatable Azure lab connecting telemetry, detection and network containment. Five controlled trials. One useful finding: detection, not the response action, dominates the wait.
Inside the research ↗One host · One scenario · Five trials
Early lab evidence; evaluation ongoing.
03 / The person behind the code

I like the point where an idea becomes a working system. The interface someone uses. The API that makes it useful. The infrastructure that keeps it running.
I’m a final-year Computer Science student at WSB Merito University, specialising in Software Development. Alongside my studies, I build full-stack products and integrate language models into real workflows.
Recent work includes a local model data pipeline, marketplace integrations and an Azure incident response lab. Each project has taught me something different about building reliable software.
Experience / 2026
Software Engineering Intern · Kraków (Remote)
Full-stack development across two B2B SaaS products, from database schema to deployment: Pro2Move and PazarPilot.
04 / Tools with a purpose
The right tool is the one that
helps the whole system work.
React · Next.js · TypeScript
Responsive product interfaces, typed API clients and end-to-end browser tests.
Python · Ollama · Qwen · Gemini
Local inference, retrieval-augmented generation and structured data extraction.
FastAPI · Fastify · PostgreSQL
APIs, authentication, data modelling and deterministic business logic.
AWS · Azure · Docker · CI/CD
Cloud storage, deployments, security automation and repeatable environments.
On my desk / Thesis research
A repeatable Azure lab connects Sentinel detection to Logic Apps containment. Initial measurements cover five controlled trials; the dedicated research chapter explains the results and their limits.
2025 Foundations Associate
Passed 7 January 2026
Credentials / 2026
Open the original certificate to see the details.
05 / The next chapter
Looking for someone who can connect the dots?
For software engineering roles or project conversations, send me a note. I’m based in Wrocław and have worked with remote teams.
A direct line
Email is the best way to reach me. Tell me what you’re working on, and let’s take it from there.
Write me an email