Jonathan Gong

Jonathan Gong

Information systems student at Carnegie Mellon.

Hi, I’m Jonathan — thanks for stopping by. I study Information Systems and Artificial Intelligence at Carnegie Mellon, though most of what I know came from building products and taking them to market: the unglamorous part where you have to get real people to use the thing. In June I helped build an offline voice assistant that lets factory operators ask questions out loud, and it somehow beat out more than 200 other teams. What keeps me interested is the seam where AI systems meet the people who have to live with them. If you’re working on something near there, or just want to say hello, my inbox is open.

Carnegie Mellon University — B.S. Information Systems, double major in Artificial Intelligence. August 2025 – May 2029.

996 Ventures — Co-Founder and CTO

  • Built a full-stack venture studio site in React, TypeScript, and TanStack Start (SSR) on Cloudflare Workers
  • Engineered a serverless submission pipeline with TanStack server functions, Zod validation, and the Google Sheets API
  • Crafted a motion-driven UI (Motion, Radix, Tailwind) with a custom intro loader and geometric component system

Corgi — AI Engineer

  • Developed an internal MCP server enabling company-wide tool integrations across teams
  • Built a React Native iOS app for Corgi Cafe, including order integrations with Square
  • Built a company-internal chatbot to streamline employee support across existing integrations

ScottyLabs — Head of Go-To-Market

  • Synthesized research from 50+ users to drive 5+ product design iterations across software projects
  • Aligned cross-functional teams of engineers, designers, and marketers around shared OKRs each sprint
  • Led launch of 3 projects, driving 1,000+ signups through organized marketing funnels

Scurry — AI Consultant

  • Developed internal tooling to streamline client onboarding, cutting turnaround time by 80%
  • Collaborated directly with clients to translate business goals into automated AI systems
  • Designed and implemented workflow automations within client HubSpot CRMs

Bizybear — VP of Communications

  • Scaled user signups from 1,000 to 60,000+ through guerilla and social media marketing
  • Translated user pain points into developer feature requests, reducing reported bugs by 30%
  • Built a company ambassador program, selecting 20+ candidates through a competitive process

ManuAI

  • Won 1st place against 200+ competitors at the Y Combinator × Moss Conversational AI Hackathon, sponsored by Unsiloed AI (YC F25)
  • Offline multimodal factory voice assistant letting operators verbally query manufacturing issues
  • Local RAG pipeline using vector embeddings to retrieve over SOPs, OEMs, and chat logs

PatientScope AI

  • Clinical decision-support platform for ICU discharges built on the MIMIC-IV dataset
  • Random Forest model predicting patient risk scores from demographic, vital, and lab features
  • Chained CTEs in PostgreSQL reconstructing and cleaning raw EHR cohorts
  • Split-hosting architecture: static frontend, FastAPI + Postgres behind an nginx reverse proxy with TLS

Repository

Chalk

  • Interactive AI voice tutor that draws synced whiteboard animations from personalized uploads
  • Designed a JSON format for the AI to describe drawings, which a separate renderer animates on screen
  • RAG pipeline (extract, chunk, embed, retrieve) grounding responses in uploaded course material

Poker Bot

  • Heads-up poker agent for a 27-card modified Hold’em tournament
  • Real-time opponent-modeling layer computing 15+ analytics to influence bot heuristics
  • Pure-Python hand evaluator with tiered ranking and a NumPy Monte Carlo equity engine running 180 simulations per decision

C0 Virtual Machine

  • Bytecode interpreter for the C0 virtual machine over a stack and heap model
  • Garbage collector managing heap-allocated memory during execution
  • Native function call support bridging interpreted bytecode to library-level operations

1st Place — Y Combinator × Moss Conversational AI Hackathon

Won against 200+ competitors with ManuAI, an offline multimodal voice assistant for factory operators.