I'm Abdul Afif Al Kaysan, a builder in Bandung, Indonesia. I build quant research systems, web3 security tooling, and Android/kernel tooling β and I publish the negative results, not just the wins.
Every claim below is checkable in the linked repository.
arbitrage-research β Measured arbitrage research across 10+ classes on real public order books, at a $100 retail basis. β The verdict is negative: no class is net-positive, because the retail fee/spread floor β not the absence of signal β defeats every one. The methodology documents three of the project's own measurement artifacts caught before they became claims.
honest-backtest β Statistical guards that refuse to certify a backtest you have fooled yourself with. β 301 tests pass in CI, each guard written from a specific false positive it caught in the author's own research.
rfc5322 β An RFC 5322-conformant email address parser (full obsolete Β§4.4 grammar), zero runtime dependencies. β 263 tests, 98% coverage, CI green on Python 3.12 and 3.13.
chainguard-web3-security β Next.js web3 security dashboard that simulates and risk-scores EVM and Solana transactions before signing, with a companion set of onchain contracts. β 8 Solidity contracts under Foundry; 91 tests passing, CI green.
dns-shield β Diagnoses ISP DNS poisoning and works around it with DNS-over-HTTPS resolution and SNI-preserving connections. β 289 tests passing; stdlib only, no runtime dependencies.
gki-abi-checker β Android GKI
ABI compatibility checker for kernel .config files and source definitions.
β Validates memory-layout flags (e.g. CONFIG_SLUB_DEBUG) whose change would
shift struct page and bootloop a GKI device. MIT.
Open to work, contract, and research collaboration in quant research, web3 security, and Android/kernel tooling.
Python Β· TypeScript / JavaScript Β· Solidity Β· Shell Β· Java Β· C β with pytest, Foundry, Node.js, GitHub Actions, and Linux/Android kernel tooling.
- Email: alkaysan07@gmail.com
- GitHub: @beduldul