Hi, my name is

BO(LEON) HE

Machine Learning Engineer · AI Applications

M.S. in Computer Science candidate at Northeastern University, graduating December 2026, with experience building scalable backend systems, ML-driven products, and XR applications. I enjoy solving complex engineering challenges, designing resilient architectures, and delivering production-ready solutions that balance performance, reliability, and maintainability. Currently seeking New Graduate Software Engineer opportunities to apply my skills in distributed systems, backend engineering, and full-stack product development.

Tech I work with

  • Python
  • PyTorch
  • LLMs / RAG
  • LangChain
  • Hugging Face
  • pgvector / FAISS
  • FastAPI
  • Ray
  • MLflow
  • Docker
  • AWS SageMaker
  • TypeScript

Featured Projects

View all →
Sweepstakes MarketingFull-Stack

Sweepstakes Marketing

Raffle module for the Marketing Platform, boosting engagement with raffles, points, rewards, redemptions, and rebate campaigns.

  • Java
  • Spring Boot
  • MyBatis
  • MySQL
Tech PiFull-Stack

Tech Pi

A full-stack developer community for sharing technical content, tutorials, and AI-powered tools.

  • Java
  • Spring Boot
  • MyBatis-Plus
  • MySQL
Dynamic AI Agent SystemAI

Dynamic AI Agent System

A dynamic AI agent platform for autonomous planning, tool execution, real-time reasoning, and workflow automation using MCP and RAG.

  • Java
  • Spring Boot
  • Spring AI
  • MyBatis
A
AI

AI Research Assistant

RAG-powered assistant that answers questions over large document corpora with cited sources.

  • Next.js
  • Python
  • FastAPI
  • LangChain
R
Backend

Realtime Chat Platform

Horizontally scalable chat backend handling 50k concurrent WebSocket connections.

  • Go
  • WebSocket
  • Redis
  • PostgreSQL
D
Full-Stack

DevTools Analytics Dashboard

Full-stack dashboard visualizing CI/CD metrics and deployment health in real time.

  • Next.js
  • TypeScript
  • tRPC
  • Prisma

About

I like taking machine learning from notebook to production — the data plumbing, serving, evaluation, and the unglamorous glue that makes a model actually useful. I gravitate toward problems where applied ML meets solid engineering. Lately that's meant building RAG systems and LLM-powered features end to end, with an eye on latency, cost, and correctness. If you're working at that intersection, I'd love to connect.