IBRAHIM BAHADIR ALTUN

PhD Student (Graph Machine Learning) at State University of New York at Buffalo

PhD student focusing on graph machine learning; developed stochastic link prediction algorithm in C++23 with sub-3 ms latency on graphs with up to 7 million events; extracted probability vectors from network streams via Bayesian inference; improved graph neural network baselines in PyTorch; served as Teaching Assistant; Web Co-Chair for ICWSM 2024; designed a workflow to run 1,000 experiments in a distributed Linux environment; left the program in good standing.

Quantitative Researcher (Cryptocurrency Trading) at TRK Technology

Instrumented a graph embedding algorithm in C++23 (55K graphs, 10M edges) with low latency; formulated a probabilistic custom loss function for LSTM training in PyTorch yielding 60%+ profit rate.

Software Engineer (Full Stack) at Toolio

Led platform service development (Minimum Display Quantity), covering technical design, GraphQL typings, Node.js engine, MySQL tables, Vue.js components; contributed to 3% of annual revenue by customer acquisition; implemented Datadog integration and dashboards; developed Conditional Formatting component and migrated to a new codebase; implemented Data Mapper and configurable mapper profiles with REST APIs.

Undergraduate Research Assistant (LLMs) at TOBB University NLP Lab

Co-authored two papers in competitive NLP/IR venues; evaluated six document selection algorithms for reliability, fairness, and reusability; developed methods to protect social media users from AI tagging (leading to 20% decrease in BERT LLM tagging); ranked 5th in Turkish NLP Competition with a text-analysis tool including stance detection.