Tsung-Yueh (Eric) Lai
NEU '27

Tsung-Yueh (Eric) Lai

AI & software engineer · Taipei → Boston

I'm a master's student in Computer Science at Northeastern University, and I came to AI and software from mechanical engineering. This unconventional path gives me a distinctive perspective, especially as robotics and AI continue to converge.

I'm currently a Software Engineer Co-op at Waterside Group in Boston, building a multi-tenant AI chatbot platform and RAG pipeline for a 28-company hospitality and real estate portfolio, and leading a small team re-architecting a React Native app into native iOS, Android, and web.

Before that I was an AI Engineer at Hyson Technology (Top 50 AI Companies in Taiwan 2025), where I led collaborations with universities, government agencies, and museums to build AI systems ranging from fraud-preventing seal recognition to underwater fish identification.

What drives me is a passion for trying everything and making it perfect. Whether it's expanding a police speed detection system or building a TOEFL practice chatbot, I'm always looking for ways to create more impactful solutions.

Résumé (PDF)Email me

Publications
3
Awards
5
Seal accuracy (gov. / company)
97% / 92%
Startup
1

Now building: Edify Entertainment, Co-founder & CEO. A game studio that edits this world: we take pieces of the real one and shape them into something we love.

When I'm not coding: Piano, Dance, Calligraphy, Photography, Gaming, Badminton.

Where I worked

Work

Jul 2026 – Present Boston, MA

Software Engineer Co-op

Waterside Group

  • Built a multi-tenant AI chatbot platform for a 28-company hospitality and real estate portfolio, shipped as a package with configurable personality and per-business knowledge bases; live on Longfellow Design Build, more brands in development.
  • Engineered an end-to-end RAG pipeline using Anthropic Claude, Voyage AI embeddings, and vector search over Payload CMS content, enabling business-specific answers from unstructured site data.
  • Integrated the HubSpot CRM API so chatbot conversations create and route contacts automatically, converting site visitors into tracked sales leads without manual entry.
  • Led a 3-person team re-architecting the Fireside field reporting app from React Native into three native platforms (iOS in Swift, Android in Kotlin, and web in Next.js), personally owning the Swift implementation and the shared architecture redesign.
  • Co-building a lead portal for Longfellow Design Build: a pipeline that verifies, enriches and filters raw property records into leads, and a web tool where the sales team calls straight from the browser.
  • Maintaining the Longfellow Design Build website: new features, bug fixes, and Lighthouse scores of 100 across Performance, Accessibility, Best Practices and SEO.

TypeScript, Next.js, Anthropic Claude API, RAG, Payload CMS, Swift, Kotlin, PostgreSQL

Jan 2024 – Aug 2025 New Taipei, Taiwan

AI Engineer

Hyson Technology Inc.

  • Led a 5-person team developing an AI seal recognition system with National Chengchi University, achieving 97% accuracy for government seals and 92% for company seals; deployed by the Kaohsiung Police Department.
  • Architected a cloud-based marine species identification platform processing 40TB of underwater video, identifying 20+ coral reef fish species with Taiwan’s National Museum of Marine Science & Technology.
  • Developed an AI traffic accident dashcam speed calculation system achieving 8.7% RMSE, reducing officer review time by 90% (30 min → 3 min); won 1st Place at the 2025 AIGO Contest (Top 4.8%, 2/41 teams) and adopted by the Kaohsiung Police Department.
  • Presented AI systems at the Kaohsiung Smart City & IoT Exhibition and the Taiwan AI Association’s 2025 AI Forum.

Python, PyTorch, OpenCV, VGG16, SuperGlue, YOLOv8

Apr 2024 – Sep 2024 New Taipei, Taiwan

AI Engineer Intern

Giant Power Technology Biomedical Corp.

  • Designed the first AI-powered TCM sleep disorder diagnosis system, integrating HRV biometric data with an LLM using RAG and Chain-of-Thought, achieving 70% accuracy on 23 cases.
  • Published as first author at the 2025 ICEIB Conference through an Industry-Academia Research Partnership.
  • Presented COMGO instrument operations and AI-assisted TCM diagnosis capabilities at medical exhibitions.

Python, LangChain, RAG, Chain-of-Thought, Gemini API

Sep 2020 – Jun 2024 New Taipei & Taipei, Taiwan

Model United Nations Instructor / Advisor

Private Nanshan Senior High School & Taipei Medical University

  • Facilitated student-led discussions on pertinent international issues and taught formal MUN conference procedure.
  • Provided comprehensive guidance on meeting protocols and revised English documents for clarity and precision.
  • Nurtured students' global awareness and honed their problem-solving skills.
  • Instructed college students at Taipei Medical University and facilitated training sessions for chairpersons.

Leadership, Public Speaking, Mentoring, English

Where I studied

Master of Science in Computer Science

Northeastern University

Sep 2025 – Aug 2027 · GPA: 3.83/4.0 · Focus on AI/ML, Software and Full-Stack Development

Bachelor of Science in Mechanical Engineering

National Taipei University of Technology

Sep 2020 – Jun 2024 · GPA: 3.86/4.0

What I've built

Projects

Multi-Tenant AI Chatbot Platform

One chatbot that each brand in a 28-company portfolio tunes for itself, answering from its own site through RAG and handing leads to HubSpot.

TypeScript, Next.js, Anthropic Claude API, RAG, Voyage AILive ›

AI Forged Seal Recognition System

Spots forged seals without real forgeries to train on: 97% accuracy for government seals and 92% for company seals. Best Paper, 2025 IEEE ICCBE.

Python, PyTorch, VGG16, SuperGlue, YOLOv8

Marine Species Identification Platform

Recognizes 20+ coral reef fish species across 40TB of underwater video, for Taiwan's National Museum of Marine Science & Technology.

Python, Computer Vision, Deep Learning, Real-time Processing

AI TCM Sleep Disorder Diagnosis

Reads heart-rate variability with an LLM to diagnose sleep disorders the TCM way: 70% accuracy on 23 cases, checked against practitioners.

Python, LangChain, Gemini API, RAG, IoMT

AI Traffic Accident Speed Calculation System

Calculates a car's speed from dashcam footage with 8.7% RMSE, cutting police review from 30 minutes to 3. 1st Place, 2025 AIGO.

Python, Computer Vision, Deep Learning, Video Processing, GCP

And the loose pages

  • LeaseQA

    A Q&A forum for lease problems that connects people straight to attorneys.

    Built with React, Next.js, Node.jsCode ›

  • Kambaz LMS

    A Canvas-like learning system: courses, assignments, grading, sign-in.

    Built with Next.js, React, TypeScriptCode ›

  • Autheria - TOEFL Speaking Bot

    My first time wiring an LLM into something: a LINE bot that coaches TOEFL speaking. Whisper listens, Gemini answers.

    Built with Python, Whisper API, Gemini APICode ›

  • ITour-AR Campus Explorer

    My first game: an AR campus tour where freshmen scan buildings to catch monsters that know the campus.

    Built with Unity, AR, C#

  • FAL Shooting Game

    Where it started: a shooting game you aim with your finger, the camera watching your hand instead of a mouse.

    Built with Python, Mediapipe, PygameCode ›

  • NSMUN News Platform

    My first website that real people used: the news site for NSMUN 2024.

    Built with HTML, CSS, JavaScriptCode ›

Pasted in

Papers and prizes

Papers

2025 ICEIB Conference

AI-Driven Diagnosis of Sleep Disorders in TCM: An Inquiry and Reasoning System with IoMT Integration

first author

2024 Conference on Investigation Technology and Forensic Science

AI Traffic Violation Detection based on Depth and Gradient Angle Change

Prizes

2025 IEEE ICCBE Conference

Best Conference Paper Award

Multi-Level Feature-Matching System for Counterfeit Seal Image Recognition

1st Place, High Distinction Award2025 AIGO Contest (Top 4.8%, 2/41 teams)

AIGO Special Excellence Award, AIGO Potential AwardTaiwan AIGO Competition 2024

Outstanding Research Paper Award2024 Forensic Science Conference

Shows & press

  • Kaohsiung Smart City & IoT ExhibitionPresented the AI Forged Seal Recognition System
  • Taiwan AI Association 2025 AI ForumUbiquitous AI Conference
  • 2024 Medical Taiwan ExhibitionPresented the COMGO instrument and AI-assisted TCM diagnosis
  • Bio-Asia Taiwan ExhibitionShowcased biomedical AI solutions
  • Press InterviewMedia coverage of the AI Seal Recognition System

What I use, and where to find me

Tools and contact

Languages
Python, Java, JavaScript, TypeScript, Swift, Kotlin, SQL, C, C++, HTML/CSS
AI/ML
PyTorch, TensorFlow, OpenCV, RAG, LLM Integration, Keras, LangChain, Anthropic Claude API, Google Gemini API, Voyage AI Embeddings, Whisper API, YOLOv8
Web & Mobile
React, React Native, Next.js, Node.js, Express.js, RESTful APIs, Full-Stack Development, Payload CMS, HubSpot
Data & Cloud
PostgreSQL, Supabase, Neon, MongoDB, Firebase, Vector Databases, Google Cloud Platform, Docker
Practices
Git, CI/CD, Agile/Scrum, Cross-Functional Leadership, Computer Vision, Deep Learning, NLP
Certificates
Building RAG Agents with LLMs (Jun 2024); Basic Theory and Practice of Deep Learning (Dec 2023); Develop AI Applications with Jetson Nano (Nov 2023)

If found, please return to
tsungyuehlai03@gmail.com

The fastest way to reach me is email. I read everything and reply to most messages within a couple of days.

GitHubLinkedInRésumé