Skip to content
View qicai14's full-sized avatar

Highlights

  • Pro

Block or report qicai14

Block user

Prevent this user from interacting with your repositories and sending you notifications. Learn more about blocking users.

You must be logged in to block users.

Content in all repositories owned by your account will be closed.
Maximum 250 characters. Please don’t include any personal information such as legal names or email addresses. This note will only be visible to you.
Report abuse

Contact GitHub support about this user’s behavior. Learn more about reporting abuse.

Report abuse
qicai14/README.md

Hi, I'm Qi Cai πŸ‘‹

Software engineering student in Malaysia. I build full-stack web and mobile apps, and I ship them with AI coding agents: I write the plan, the rules and the tests, and the agents do the typing under review.

  • πŸ”­ Building NotaLah, my final year project: a peer-to-peer study-notes platform with an explainable ranking feed
  • πŸ† Finalist, MUM x Averis AI Hackathon 2026 with a multi-agent shipping-document verifier
  • πŸ€– Agentic engineering: Claude Code with project rules, hooks, git worktrees and review loops
  • πŸ“« qicailim1114@gmail.com

Languages

TypeScript JavaScript Python Java C++ Dart PHP R SQL HTML5 CSS

Frameworks and tools

React Vite Tailwind CSS Node.js Express FastAPI Socket.io Flutter PostgreSQL SQL Server MongoDB Docker GitHub Actions Cloudflare Vercel Render

AI and agents

Claude Code GitHub Copilot Antigravity Google ADK Gemini Qdrant RAGAS

Featured projects

NotaLah: final year project, solo. A note-sharing platform for university students: slide editor, ranked feed, real-time chat and collaboration, lecturer verification and moderation. React 19, Express 5, PostgreSQL 17, Socket.io, Cloudflare R2. ~38k lines of TypeScript, 570+ tests, 50+ reviewed PRs, CI on every push. Built with an agentic workflow: CLAUDE.md rules per package, 75 logged design decisions, feature specs, session handovers, isolated worktrees with their own test DB, and a review pass on every PR.

La Peace SDOC: finalist, MUM x Averis AI Hackathon 2026 (team of 4). Multi-agent verification of shipping documents: Google ADK agents on Gemini, Qdrant RAG, OCR extraction, RAGAS evaluation, FastAPI + MongoDB. I designed the extraction pipeline: deterministic OCR first, escalating to a vision LLM only when needed, to keep runs fast and cheap. I also built the React/Vite frontend (verification workspace, AI copilot drawer, summaries), fixed backend features alongside the UI, and shipped the Vercel deployment. Live demo

SME FundMatch: AWS Great AI Hackathon 2025, team MVP matching small businesses with funding. Live demo

Coursework

Pinned Loading

  1. NotaLah NotaLah Public

    Development of a Peer-to-Peer Academic Resource Platform with Adaptive Ranking Algorithm

    TypeScript

  2. jaga jaga Public

    Android scam-call guardian for Malaysian families. Muse Voice Transcribe hears the live call, free rules track the scam script, and Muse Spark judges only when they fire. It warns the protected use…

    Dart 1

  3. ambatuCode887/la-peace-multi-agentic ambatuCode887/la-peace-multi-agentic Public

    TypeScript

  4. TeddyHuZz/AWS-Great-AI-Hackathon-2025 TeddyHuZz/AWS-Great-AI-Hackathon-2025 Public

    JavaScript