Most of Web3's hard problems aren't cryptographic — they're about deciding who's allowed to do what, and proving it after the fact. That's the thread through everything below.
Where I started — Full-stack Web3, hackathon hacker. Ship it, prove it works, move on. BNB Chain Quarterfinals, wins at EduChain and Tezos EVM Hack.
Where I found the real problem — Building T-REX (ERC-3643) compliance infra at TheOpenAssets across four chains, I realized "compliant" isn't a checkbox, it's a system that has to hold state correctly under async execution. Most teams fake this. I got interested in not faking it.
Where I built for real — Founding engineer at PayZoll: payroll infra with real settlement, real money, zero tolerance for "mostly works." Rust/Soroban on Stellar, EVM on the side.
Where I am now — Building an agentic policy layer: permissioning, authorization, and audit trail for autonomous AI agents transacting on-chain. Also building a payment gateway from scratch. Same instinct as always — an agent (or a person) should only be able to do what it's actually allowed to do, and there should be a trail proving it did.
Background:
- Founding Engineer, PayZoll (Oct 2024 – Apr 2026) — payroll & cross-border payments infra
- Smart Contract & Infrastructure Engineer, TheOpenAssets/Credora (Sep 2025 – Feb 2026) — RWA compliance infra
- Winner, Zera (EduChain) and SlashMarket (Tezos EVM Hack), plus BNB Chain Quarterfinals
- Currently: agentic policy layer + a payment gateway build, both in progress
Most of the interesting parts are architecture decisions more than repos — happy to talk through any of it directly.