~/manseung $ whoami
Hi, I'm Brendan Manseung Choi. An AI/full-stack engineer who takes products from architecture to production — currently building and deploying a clinical AI system used by real patients.
I'm a Computer Science graduate from the University of Auckland with First Class Honours (GPA 8/9), specializing in AI and software development. My journey in tech has been driven by a genuine passion for creating innovative solutions using the latest tools and technologies.
I currently serve as Founding Engineer & CTO at Align Health Tech, where I take AI-powered products from architecture to production—including a clinical AI system now live with real patients at an Urgent Care Clinic.
Beyond coding, I'm also a certified personal trainer (NASM), electric guitarist, and fitness enthusiast. These diverse interests help me maintain a balanced perspective and approach to problem-solving.
A selection of my work in AI, fullstack, and applied software development
Typing practice for programmers, inspired by Monkeytype but focused on code.
• Improves coding fluency by typing realistic code snippets
• Designed for speed + accuracy with programming-focused tests
Productivity mobile app (React Native + Expo) designed around the Pomodoro technique.
• Implemented Stripe IAP + RevenueCatfor subscription & trial management
• Released iOS beta and onboarded 20+ testers with retention feedback
• Built scalable Firebase backend and adaptive UI for iOS/Android
Lightweight VS Code / Cursor extension to boost developer workflow.
• Enables quick file prioritizationwith a "star" marker
• Published on the VS Code Marketplace
• Built in TypeScript with minimal overhead for adoption
On-device AI coding assistant using Ollama API.
• Published on the VS Code Marketplace
• Integrated local LLMs for privacy-first developer support
• Easy to use and install, open source, and free to use
🏆 1st Place out of 12 teams at the University of Auckland Legal Tech Hackathon 2025.
• Built a RAG pipeline translating complex legal contracts into verified plain-English summaries
• Ensured accuracy + personalized support using RAG pipelines
• Integrated Tinfoil API to guarantee privacy-preserving AI usage
• Demoed to judges with positive feedback for real-world impact in legal accessibility
Hackathon project addressing sustainability with blockchain.
• Designed tokenized incentives for carbon offsets
• Built smart contracts on Ethereum Sepolia with Solidity
• Delivered prototype + pitch deck in under 48 hours with a team of 4
Research investigating privacy-preserving, offline-capable clinical decision support for NZ healthcare.
• Built an on-device RAG system with quantized SmolLM2 models (135M–360M parameters)
• Achieved 70.4% triage accuracy within 300 MB memory and 2.4s end-to-end latency on a retail consumer device (iPhone 14 Pro)
• Trained 4 LoRA adapter configurations on 13,825 synthetic triage cases
• Demonstrated 30–50× fewer parameters than existing compact medical models
• Established foundations for equitable, offline-capable AI-assisted healthcare
Prototype of an LLM-powered crypto trading agent.
• Built with Flask + Web3.py + Gemini API
• Enabled users to execute trades in plain English(e.g., "Buy 0.1 ETH if BTC rises 2%")
• Produced demo paper & video documenting feasibility of agentic finance workflows
AI-powered course assistant, awarded 2nd Placein University of Auckland's 2024 COMP SCI 399 capstone.
• Built with GPT-based NLP, TTS voice synthesis, and an avatar-based React frontend
• Designed conversational interfaces to replace static course information pages
• Delivered in agile sprints with GitHub + JIRA; continued development as a Research Assistant
A comprehensive look at my technical expertise — with clear core strengths up front
My professional journey and academic background
December 2025 - Present
Building an end-to-end AI voice agent for inbound calls, lead qualification, and automated job booking — VAPI and Twilio for call agents, n8n for workflow automation, GoHighLevel for CRM/scheduling sync, and Stripe for subscriptions. Sold a live subscription to a real user.
October 2025 - July 2026
Designed and implemented AI-driven clinical workflow automation for urgent care clinics, built as a Next.js / TypeScript application on Azure. Secured NZD $15K Velocity Challenge funding. Ongoing pilot at an Urgent Care Clinic with an average patient satisfaction score of 8/10 across 120 pilot patients.
University of Auckland
December 2024 - 2025
Built an AI-powered course assistant using GPT-based NLP, TTS voice synthesis, and an avatar-based React frontend. Designed conversational interfaces to replace static course information pages.
University of Auckland
February 2022 - December 2025
Studied Machine Learning, AI, and large-scale software development with a GPA of 8.063/9. Graduated with First Class Honours and awarded the University of Auckland Postgraduate Honours/PG Diploma Scholarship.
Self-Employed
2026 - Present
Teaching students maths and programming through a learning platform focused on clear explanations, problem solving, and practical coding skills.
CityFitness
September 2024 - 2025
NASM certified personal trainer, helping clients achieve fitness goals through personalized workout programs and nutrition guidance.
Self-Employed
May 2023 - 2024
Provided one-on-one mathematics tutoring to high school and university students, focusing on calculus and algebra.
Coin Singing Booth Karaoke
May 2022 - 2025
Managed customer transactions and provided excellent customer service in a fast-paced entertainment environment.
Katsubi
2019 - 2021
Supported restaurant operations through food service and kitchen assistance, developed strong teamwork and time management skills.
My hobbies and passions