AI Engineer · jingh

Jing Huang

Machine learning. Computer vision. AI agents.

Specializing in machine learning, deep learning, and AI system architecture. 6+ years building production systems at the intersection of research and engineering.

Expertise

Skills

Machine Learning & Deep Learning

End-to-end model development from research to production. Computer vision, time-series forecasting, and multimodal systems.

TensorFlowPyTorchScikit-learnCNNLSTMTransformer

AI Agents

Designing and deploying multi-agent systems. Proficient with Google ADK and LangChain/LangGraph for orchestration, tool use, and retrieval-augmented generation.

Google ADKLangChainLangGraphRAGTool Use

Computer Vision

OpenCVMediaPipeYOLOsTensorRT

Data Analysis

SQLPandasMatplotlib

Python

NumPyDjangoOpenCV

C / C++

C++STM32Qt

Backend

DjangoFlaskFastAPI

Frontend

ReactNext.jsTypeScript

Cloud

AWSGCP

Work

Projects

Currently building

Poiesis AI

A boutique-style AI agent platform. Pre-built agents accessible via web interface, or integrated directly into Telegram and Feishu bots — enabling cross-platform AI workflows without infrastructure overhead.

LangChainLangGraphTelegram Bot APIFeishuNext.jsFastAPI
Visit platform →
  • Curated library of pre-built agents
  • Web interface for direct access
  • IM channels integration
  • Cross-platform AI workflows

Surgical Video Analytics Platform

  • Object Detection & Tracking
  • Human Pose Estimation
  • Speech-to-Text
  • OpenAI Integration
PythonPyTorchTensorRTAWS

Surgical Video Analytics Platform

Built a web platform that analyzes surgical videos using multiple AI modules in parallel, and integrates with OpenAI for automated report generation.

View project →

On-device Instrument Detection

  • YOLO-based detection on iOS
  • CoreML & Unity deployment
  • 98%+ recognition accuracy
UnityCoreMLONNXXcode

On-device Instrument Detection

Developed an object detection model for surgical instruments and deployed it to a mobile application using Unity and Xcode.

View project →

Ultralytics YOLOs GUI

  • Qt GUI App
  • Object Detection & Tracking
  • Human Pose Estimation
  • Instance Segmentation
PythonPyQt5OpenCVYOLO

Ultralytics YOLOs GUI

Built an open source GUI for object detection, segmentation, and tracking using Ultralytics YOLOs on video, image, and live stream sources. All configurable through a clean operator UI without requiring any coding knowledge.

View project →

Battery State Prediction

  • Research Project
  • Data Engineering
  • Time-series forecasting on STM32
PythonTensorFlowCSTM32MatLab

Battery State Prediction

Trained a time-series model to predict battery state of charge and health by analyzing volume changes in cells, quantized and deployed it to an STM32 microcontroller.

View project →

Remote Tower System for ATC

  • Real-time panoramic surveillance
  • Multi-camera aircraft detection & tracking
  • C++ streaming APIs
PythonC++TensorFlowCVATAWS

Remote Tower System for ATC

Developed a remote tower solution for air traffic control using AI and computer vision technologies. The system provides air traffic controllers with a real-time panoramic video feed of the airport environment, complemented by automated object detection and tracking. This enhances safety and efficiency of air traffic.

View project →

Sales Forecasting Web App

  • Time-series demand forecasting
  • RESTful backend development
PythonFlaskTensorFlowGCPMySQL

Sales Forecasting Web App

Developed a web application for retail sales forecasting using AI to optimize supply chain management. Researched forecasting algorithms and implemented backend for model training, evaluation, and inference.

View project →