Behzat Aslanoğlu
از مولکولها تا نورونها تا agentها
ریاضیات و کد با هم در ذهن من رشد کردند — سپس از مولکولها، مغزها و یادگیری ماشین گذشتند تا به مرزهای پیشرو رسیدند: LLMها، هوش مصنوعی agentic و ازدحام agentها. یک رشتهٔ پیوسته.
قوس موازی.
Math never waited for the code, and the code never waited for the math. They arrived together and stayed together — one continuous, intertwined thread from the very start.
— the through-line · BSc → MSc → PhD → the work
این نظم هرگز درباره حفظ کردن فرمولها نبود — درباره تجزیه یک مسئله واقعی و آشفته به معادلاتی بود که واقعاً میتوانید حل کنید.
— applied mathematics · University of Tabriz
Don't just fit the curve — model the mechanism.
— molecular dynamics · TÜBİTAK, F876L resistance story
Vision
Intelligence from molecules to neurons to agents — one unbroken thread. Mathematics, science, and agentic AI grown together, and shared as I build.
Mission
Build LLMs and agentic AI that don't just answer — they act. From first-principles ML to coordinated swarms of agents, ship systems, not notebooks.
Philosophy
Reduce a messy, real system to a model you can trust. Respect the physics, model the mechanism — don't just fit the curve. Math and code arrived together and never waited for each other.
The Goal
Coordinated swarms where emergence echoes the brains I studied — resilience, parallel exploration, and answers that come from collective reasoning rather than one model's guess.
سه فصل، یک رشته
ستونهایی که روایت بر آنها میچرخد — و موتوری که هنوز با آن میسازم.
From Proofs to Physics
Applied mathematics, molecular dynamics, and the F876L resistance story. The discipline of decomposing a messy, real problem into equations you can actually solve.
Connectomics as Graph Theory
Super-resolution for the most complex known graph — predicting high-resolution brain graphs from sparse, noisy observations. Open data, released to the community.
From ML to Swarms
From-scratch ML → federated learning → agents. Today: LLM systems end to end, and a working thesis that emergence echoes the brains I studied.