Skip to content

@weixinleow

Weixin Leow

@weixinleow

React
Skill iconGit
claude
copilot
hermes

Shah Alam, Malaysia

Hi, I'm Leow Wei Xin

Software Engineering undergrad @ Universiti Malaya ยท Tan Sri Teh Hong Piow Legacy (Public Bank) Scholar
๐Ÿ“ง [email protected] ยท ๐Ÿ”— LinkedIn ยท ๐Ÿ’ป GitHub

Applying to build at MUBA Blockchain Hackathon 2026.


About Me

I'm a Computer Science (Software Engineering) student who ships full products under pressure, not just prototypes. In the last year I've led or co-led four hackathon teams โ€” winning an AI award, a 1st Runner-Up finish, and a Top-6 placing โ€” while balancing a full course load and a scholarship. I move fast between roles: product manager one weekend, backend architect the next, frontend engineer after that. That range is what I want to bring to a blockchain build.

I'm currently an IT Intern at Etiqa Insurance and Takaful, where I get a front-row view of how a regulated financial institution thinks about trust, risk, and transaction integrity โ€” exactly the problems blockchain is built to solve.

What I'm Good At

  • Full-stack product delivery โ€” React Native, Next.js, NestJS, Node.js, Java, C++, PostgreSQL, Redis, and Streamlit/Python across four shipped projects
  • AI-agent architecture โ€” designing multi-agent systems (forecasting, risk-scoring, negotiation agents) that make autonomous decisions, not just chat
  • Trust & risk systems โ€” building escrow flows, fraud/risk scoring, and dispute-resolution logic for fintech and insurance use cases
  • Team leadership under a clock โ€” leading teams of up to 4 through 24โ€“48 hour build cycles with zero scope creep and on-time delivery
  • Rapid learning โ€” picking up an unfamiliar stack (a new SDK, model, or protocol) fast enough to ship a working demo in a weekend

Projects I've Built

NGAM โ€” 1st Runner-Up, Devin ร— Claw Collective ร— Qwen Hackathon 2026

An autonomous SME procurement agent for Malaysian F&B businesses. A tri-modal agent architecture (Hermes orchestrator + Qwen reasoning engine + Streamlit dashboard) forecasts demand, screens supplier fraud risk, negotiates with vendors in three languages, and issues signed purchase orders โ€” end-to-end from a single Telegram message in under 90 seconds. It also runs a "Price Shock Index" that flags cost spikes before they hit supplier quotes.
Telegram Bot ยท Hermes Agent ยท Qwen ยท Streamlit (Python) โ€” github.com/xinweileow/ngam

ClariFi โ€” Best AI Award, Borneo HackWknd 2026

A voice- and receipt-based expense capture app with community-powered grocery price comparison, built as Project Manager and shipped against 15 competing teams.
React Native ยท NestJS ยท Gemini 3.1 Flash Lite โ€” github.com/FlyingShuriken/clari-fi

GigHub โ€” Top 6 Finalist, UMHackathon 2026

A full-stack freelance marketplace with milestone-locked escrow and AI-assisted brief validation โ€” led a team of 4 to build a responsive multi-role UI, milestone dashboards, and an automated 72-hour payment release flow. The escrow and time-locked payment-release logic here maps directly onto smart-contract design patterns.
Next.js ยท Node.js ยท PostgreSQL ยท Redis ยท Cloudflare R2 โ€” github.com/CookingMouse/GigHub

ClaimLin โ€” ongoing

An AI-powered insurance claims advocate helping Malaysian property owners navigate post-disaster claims: analysing policies, flagging hidden risks, calculating fair payouts, and generating dispute letters. Insurance claims are a textbook case for on-chain verifiability โ€” provably fair payouts, tamper-proof policy records, and transparent dispute trails.
github.com/CookingMouse/ClaimLin

Why MUBA Blockchain Hackathon 2026

What I bring to the table

  • A proven track record of finishing what I start: 4/4 hackathons shipped a working, judged demo, not a slide deck
  • Direct, transferable experience with the exact primitives blockchain hackathons reward โ€” escrow, milestone-based payment release, fraud/risk scoring, and dispute resolution โ€” from GigHub and ClaimLin
  • Applied AI-agent experience (NGAM) I want to fuse with on-chain infrastructure: autonomous agents that don't just recommend a transaction, but execute and settle it trustlessly on-chain
  • Leadership that keeps a team scoped and shipping inside a hard deadline, proven across four separate events

What I want out of it

  • My first hands-on build with smart contracts and on-chain infrastructure, backed by a team that already knows how to execute under a hackathon clock
  • To test whether the escrow and trust-layer patterns I've already built off-chain (GigHub, ClaimLin) hold up โ€” and get sharper โ€” when rebuilt on-chain
  • To learn directly from a blockchain-native community and mentor pool I don't otherwise have access to as a second-year CS student
  • A shot at building something that could genuinely carry for