Building infrastructure where privacy is a guarantee, not a policy. Researcher. Engineer. Writer. Someone who asks too many questions to ever stop building.
Hardware-attested, constant-time AI inference inside Trusted Execution Environments. CIR eliminates timing side-channels and produces cryptographic proof of execution integrity — clients verify how their data was processed, not just trust that it was.
Smart contract audit portfolio spanning solo practice and contest submissions. Focused on reentrancy vulnerabilities, access control failures, and economic attack vectors across DeFi protocols.
ML-based anomaly detection on Ethereum mainnet transaction data. Tree-based classifiers surface suspicious on-chain activity patterns that rule-based systems miss.
A full-stack web app for creating, sharing, and taking interactive quizzes. Built with Next.js and Node/Express — supports quiz authoring, shareable links, real-time scoring, and a personal quiz library.
I'm a computer science researcher based in Lagos, building at the intersection of confidential computing and AI infrastructure. My current work — CIR — addresses a problem most AI companies haven't admitted exists yet: the inability to cryptographically prove how a model processed private data.
I came to this problem the way most good problems find you — sideways. Web3 security taught me that trust is a technical property, not a policy one. Dostoevsky taught me that the questions worth asking don't resolve cleanly. Both turned out useful.
I write and make videos about technology, faith, and the intellectual life — not separately, but as the same inquiry from different angles.
Open to contracts, research collaborations, and consulting in confidential computing and AI security. If you're building something that needs to be provably private — let's talk.