_
My curiosity about software development keeps growing.
From the first idea to the final product, I'm captivated by the process of bringing things
to life.
At Calibre Tour, I work with .NET, PostgreSQL, Express.js, MySQL, and Next.js to build
robust systems across both frontend and backend.
On my own, I built LIBRAIN-an open-source multi-agent RAG system in .NET 10 that
ingests scientific papers, embeds them in a vector store, and generates citation-grounded
hypotheses through a Reader → Synthesis → Evaluator pipeline.
A parallel Discovery Mode bridges across a 41-paper cross-domain corpus (AI, drug discovery,
climate, neuroscience), explicitly tagging the speculative portion as a novelClaim.
The companion paper is published on Zenodo with a DOI.
My focus: clean backend systems where .NET and AI tie together.
Skills
// what I build with
html
css
javascript
react
next
vue
c#
.net
typescript
node
express
php
laravel
flutter
unity
mysql
mongodb
postgresql
mssql
kafka
rabbitmq
Journey
My personal journeyProjects
Some projects I've worked onCertifications
My achievements and credentials
Back-end Web Developer
AIESEC Global Talent
EF SET English Certificate
EF Standard English TestPublications
My research and academic contributionsLIBRAIN: A Multi-Agent RAG System for Scientific Discovery
An open-source, auditable multi-agent retrieval-augmented generation system for scientific hypothesis discovery, built end to end in .NET 10. A structural citation-validation contract drops any reference that does not resolve to the retrieved set, and a Discovery Mode fences the speculative part of each hypothesis into a separately scored novelClaim field. Every output is traceable by a correlation ID.
Mobile Application for Hearing Impaired Communication
Designed a mobile application using Unity, C#, and Speech-to-Text technologies to bridge the communication gap between hearing-impaired individuals and the general public by translating spoken language into Turkish Sign Language animations in real time.