At Token Financial Technologies, I built an end-to-end AI pipeline and backend for a merchant news-curation module planned for Token App. It turns articles from varied sources into editor-reviewable newsletter drafts using AWS Bedrock and DynamoDB.
At ROMER, I develop data-driven methods for nonverbal behavior in collaborative human-robot interaction. With a three-person research team, I created an interaction-video dataset and developed a framework that generates behavioral sequences and motion trajectories for robots.
On ALAZ, I formed and led a 13-member team developing an autonomous vehicle from scratch. We placed 4th in the closed category at the RDW Self Driving Challenge 2026, and I developed localization, perception, planning, and control modules integrated with Autoware.
On AMBALE, my team placed 1st among 358 applicant teams in the TEKNOFEST 2026 AI-Assisted Airline Optimization Competition; our graph-based decomposition reduced optimization runtime from approximately 90 minutes to 3 minutes. I also developed a real-time UAV detection and tracking system for ANZU and contributed to building the Google Developer Groups on Campus METU community and its Productive Team.
