Aman Zainal
AI Engineer · Singapore
Aman Zainal is a Singapore-based AI engineer and the founding AI engineer at AIRAP. He studies Computer Science and Mechanical Engineering at the National University of Singapore as an ASEAN Scholar (Ministry of Education, Singapore). He builds production AI products that survive past the demo — aviation compliance audit tools, B2B operations platforms, research interpretability layers, and generative AI pipelines on a local 4-GPU rig.
Selected work
- Aero Plus AI Suite (AIRAP) — aviation compliance audit; ingests ~10GB PDF/image packages and auto-fills 26-item engine and 72-item aircraft airworthiness checklists with page-level citations. ~95% accuracy vs analyst ground truth.
- EASI App (AIRAP / Epico) — B2B alcohol-distribution ordering platform; React Native mobile, four Vite/React portals, C#/.NET AutoCount bridge, Supabase. Shipped in two weeks by a two-engineer team.
- Forecast Interpreter (A*STAR ARTC) — conversational interpretability layer over A*STAR's internal demand-forecasting models. Cuts planner workflow from ~40 minutes to ~2 minutes per forecast.
- Banyan — collaborative family-tree app with realtime graph rendering and share-link onboarding. Next.js 16, Supabase Realtime + RLS, React Flow.
- Virtual Wardrobe — generative AI clothing try-on platform; FastAPI gateway, Vite/React web, Expo React Native, ComfyUI on RunPod GPU.
- Real or AI eye test — twelve-plate guessing game; half public-domain photos, half generated locally on the 4-GPU rig.
Site map
- Projects — production systems shipped with real users
- About — background, capabilities, current role
- Real or AI — eye-test game powered by local diffusion models
- Stack — the image-pipeline rig
- Contact — Telegram, email, GitHub, LinkedIn
- Resume (PDF)
Elsewhere
Find Aman Zainal on GitHub, LinkedIn, and Telegram. Email: aman@u.nus.edu.