Resume

Jason Chan

I build software at the intersection of machine learning and systems engineering.

Boston, MA — Northeastern ’27

Scroll
01 About

I’m drawn to problems where the model and the system have to work together — where training a network is only half the job.

I’ve shipped software for defense robots, won hackathons, and organized events that brought hundreds of engineers into the same room. I care about depth over breadth, and building things that actually hold up.

Reference from my manager at Teledyne FLIR →
02 Selected Work
HackMIT — Top 15 of 500+

Cosign

Problem

ASL speakers lack real-time digital tools that understand motion-based signs — most solutions only handle static fingerspelling.

Approach

TensorFlow CNN+LSTM pipeline classifying 250+ gestures from 90,000+ video frames with sub-200ms end-to-end inference, streamed via Flask REST APIs and SocketIO to a React frontend.

Result

Top 15 of 500+ teams at HackMIT. Request batching and async inference handling sustained throughput during high-volume streaming.

TensorFlow Flask React SocketIO Python
X Developer Challenge — Selected from 1,000+

Scam Detection Tool

Problem

Users have no real-time way to identify scam posts while scrolling their timeline.

Approach

Modular LLM-based system with structured Python services, automated evaluation pipelines, and REST integration with retry logic to reduce false positives through targeted model tuning.

Result

Selected from 1,000+ submissions. Automated testing improved end-to-end reliability across edge cases.

Python LLMs Few-Shot Learning X API
Best Covid-19 Hack Winner

Mask Verified

Problem

Public spaces needed automated mask compliance monitoring without manual oversight.

Approach

Trained a CNN to detect mask usage from facial landmarks in real time using TensorFlow and OpenCV.

Result

Won Best Covid-19 Hack. Achieved real-time inference on live video streams.

TensorFlow OpenCV Python

Tesla Sentiment Analysis

Problem

Understanding brand perception requires processing thousands of unstructured social posts.

Approach

NLP pipeline with VADER sentiment analysis and custom cleaning — stopword removal, lemmatization — to surface patterns.

Result

Processed 10,000+ tweets, extracting actionable brand sentiment data.

Python NLTK VADER NLP
03 Experience
Jan – Jun 2025

Software Engineering Intern

Teledyne FLIR — Unmanned Ground Systems

Shipped 8+ production features across a 100K+ LOC Android codebase — integrating real-time telemetry, 3D mapping, and live video for 4 autonomous ground robots, while hunting down 40+ threading and memory-leak bugs that were freezing the operator UI.

Nov 2023 – Mar 2024

Executive Board Member

PawHacks

Built Northeastern Oakland’s first hackathon from scratch — wrote the full-stack platform (React, Node.js, PostgreSQL) that handled 300+ registrants and 150+ concurrent users on event day, while cold-emailing sponsors until we hit $15,000+.

Aug – Dec 2022

Student Ambassador

Replit

Selected as 1 of 25 worldwide — taught 100+ students to code and beta tested Ghostwriter AI months before public launch.

04 Skills

ML & Data

  • Python
  • TensorFlow
  • PyTorch
  • NumPy
  • pandas
  • NLTK
  • OpenCV

Full-Stack

  • React
  • Next.js
  • Node.js
  • TypeScript
  • Flask
  • PostgreSQL
  • MongoDB

Systems & Infra

  • Java
  • C++
  • Docker
  • Kubernetes
  • Linux
  • Git

Exploring

  • Rust
  • Go
05 Contact

Let’s build something.

Currently open to internships and new grad roles for 2027.

chan.jaso@northeastern.edu LinkedIn ↗ Resume ↗