Hee Thung Han

I'm a software engineer and technical architect based in Singapore with over a decade of experience building scalable distributed systems. Currently at Ninja Van, I head the Technical Architecture Committee, governing decisions across 4 engineering labs and 100+ engineers while driving the organization's AI adoption initiatives.

My journey has taken me from Silicon Valley — where I built interactive medical software at Stanford's School of Medicine and prototyped computer vision systems at startups — to Southeast Asia's tech scene. Over nearly 10 years at Ninja Van, I've grown from software engineer to chief architect, along the way refactoring monolithic services handling 100M+ daily requests, building high-throughput event pipelines, and deploying the company's first data and ML systems.

These days, I'm most excited about the intersection of AI and software engineering — building multi-agent AI platforms, RAG-powered knowledge systems, and multimodal AI applications that solve real operational problems at scale. I hold an MS from Stanford and a BS from the University of Illinois at Urbana-Champaign, both in Electrical Engineering.

2024 — Present

Chief Technical Architect · Ninja Van

Managing the Technical Architecture Committee to govern design decisions and align technical strategies with business goals. Driving organization-wide AI adoption by leading exploration, training, and integration initiatives, including the implementation of multi-agent AI systems with Telegram bot integration using RAG and knowledge graphs.

AI AgentsRAGPythonSystem ArchitectureTechnical Leadership

2021 — 2024

Senior Staff Software Engineer · Ninja Van

Integrated machine learning algorithms to achieve 99.3% geocoding accuracy and standardized microservice templates to accelerate deployment. Refactored critical monolithic services, managed zero-downtime framework upgrades, and launched data archiving initiatives with Kubernetes CronJobs to reduce costs. Automated API client generation using OpenAPI specs to streamline development workflows across teams.

GoKubernetesRefactoringMachine LearningOpenAPIDistributed Systems

2017 — 2021

Staff Software Engineer · Ninja Van

Enhanced fault tolerance with circuit breakers and Kafka retry processing while establishing the Engineering Excellence team to define coding standards and upgrade CI/CD pipelines. Deployed the company's first data pipeline using Airflow and prototyped a parcel dimensioning system using OpenCV and Kinect-like cameras. Built high-throughput microservices handling over 700 Kafka events per second.

JavaKafkaAirflowCI/CDOpenCVEngineering Management

2016 — 2017

Senior Software Engineer · Ninja Van

Developed Play Java microservices for routing deliveries with Hazelcast caching to enhance performance. Led frontend projects, facilitated knowledge sharing across teams, and mentored engineers to foster professional growth.

JavaPlay FrameworkHazelcastFrontend DevelopmentMentoring

2015 — 2016

Software Engineer · Ninja Van

Redesigned the frontend architecture for a multi-country operations dashboard and spearheaded the adoption of AngularJS. Reduced deployment costs by optimizing the hosting of static web assets on AWS S3 with CloudFront CDN.

AngularJSAWS S3CloudFrontFrontend ArchitectureJavaScript

2013 — 2015

Software Engineer · Mobikon Asia

Deployed a Play Scala application to Heroku for restaurant e-menus and marketing tools. Built cross-platform AngularJS applications with Apache Cordova for mobile and web, serving restaurant chains across Singapore and India.

ScalaPlay FrameworkHerokuAngularJSApache Cordova

2012 — 2013

Programmer · Stanford University

Created interactive teaching software in Unity3D to accelerate content creation and optimize performance. Delivered specialized software for research studies involving video capture, stereoscopy, and haptics for the School of Medicine.

Unity3DC#Augmented RealityHaptics

JUN — NOV 2012

Research Engineer · Gen-9, Inc

Built home therapy applications using Kinect with real-time object tracking in C++. Designed and implemented multi-camera systems for real-time image acquisition and processing.

C++Kinect SDKComputer VisionReal-time Systems

JUN — SEP 2011

Research Intern · Gen-9, Inc

Evaluated MEMS sensor performance using Matlab to assess the feasibility of accurate indoor localization and human motion classification with machine learning techniques.

MatlabMachine LearningSensorsData Analysis

2008 — 2009

Teaching Assistant · University of Illinois at Urbana-Champaign

Designed programming assignments in Assembly and C, assisted students with debugging during office hours, and graded coursework and exams.

CAssemblyTeachingDebugging

JUN — JUL 2008

Research Intern · Infocomm Development Authority of Singapore

Researched cloud and distributed computing technologies and proposed applications for Singapore's national IT infrastructure.

Cloud ComputingDistributed SystemsResearchInfrastructure

View Full Résumé

AI Dispatcher Platform for Delivery Drivers

Nulla vel nulla est. Cras sed enim ac turpis bibendum iaculis nec facilisis sapien. Integer vel dolor ac magna laoreet volutpat. Morbi consequat cursus tincidunt. In sodales, urna eu vulputate faucibus, nulla tortor porttitor lorem, in auctor lacus.

FastAPINext.jsLangChainTelegram

Parcel Dimensioning Prototype

Built an automated parcel dimensioning system using depth cameras and PCA-based oriented bounding box fitting. Deployed across warehouses in three countries for real-time package measurement on weighing scales.

C++OpenCVPCLDepth CamerasSocket.IO

View All Projects