Behkish Nassirzadeh, Ph.D.
Entrepreneur · Blockchain Innovator · AI Researcher · AdTech Advocate
bnassirz@uwaterloo.ca | LinkedIn | GitHub | DBLP | Google Scholar | Resume
Entrepreneur · Blockchain Innovator · AI Researcher · AdTech Advocate
bnassirz@uwaterloo.ca | LinkedIn | GitHub | DBLP | Google Scholar | Resume
I hold a PhD in Computer Engineering (2025) from the University of Waterloo, co-supervised by Prof. Vijay Ganesh and Prof. Anwar Hasan. My research integrates blockchain security, AI-driven automated program repair, and decentralized advertising systems. I’ve developed tools like GasGauge, GasGuard, and GasGaugeAI, alongside decentralized protocols like CountChain and AdChain.
With over a decade of academic, technical, and leadership experience, I specialize in building end-to-end systems that are secure, scalable, and impactful. My expertise spans software architecture, decentralized consensus, security tooling, and AI-assisted automation.
Beyond research, I’m an entrepreneur passionate about technologies that challenge norms and create measurable impact. I also advocate for mental health and enjoy building solutions that bridge technical depth with human-centered value.
High-signal conversations that move the needle:
A static/dynamic analysis tool that detects high-risk loops susceptible to out-of-gas failures. It uses white-box fuzzing for precise vulnerability localization and run-time verification to find safe boundaries.
An LLM-based tool that injects dynamic safeguards to prevent gas exhaustion using a fine-tuned GPT model. It computes safe loop bounds and enforces them at execution time using require() instrumentation.
An AI-powered framework that automatically repairs gas-related vulnerabilities in Ethereum smart contracts. Combines static analysis, Foundry-based fuzzing, and multi-agent LLM feedback loops for secure code patching.
A decentralized oracle protocol for counting systems using Sybil-resistant, incentive-compatible consensus. Optimized for high-volume, real-time environments.
A decentralized advertising protocol leveraging CountChain for verifiable impressions and a reputation layer to deter fraud. Designed for programmatic ads, header bidding, and decentralized ad exchanges.