Work

Open-source background coding agent with 1.2k stars on GitHub. Feature-filled agent that works in a MicroVM with full codebase understanding.

Multimodal agent and long-context video understanding to help hollywood editors. Worked on the the most powerful video retrieval and editing agent.

A decentralized cross-device model training system with model and tensor parallelism to reduce compute needed to train large models.

Leading behaviour and interaction software for a humanoid robot design team in Waterloo. Built GPU-optimized 3D voxel grids for awareness models.

City simulation of Los Angeles with AI Agents, simulating human behaviour and optimizing transit routing with RL.

Built an autonomous model tank that navigates campus walking paths and delivers small parcels. Implemented path planning, simulations, object detection, PID control, and CV.

Built an entirely self-sustaining startup on AI Agents, which built the startup from scratch, made business plans, did branding, built a full-stack app and put a job posting on LinkedIn.

Research at Cohere for AI to represent different languages as different modalities for training multilingual language models.

Deep learning pipeline that analyzes seismic frequencies and local policy to design affordable earthquake-resistant buildings.

An offline mesh network written in Swift to allow for cross-device transfer of files entirely offline, creating a chain of encrypted nodes.

Helping with the software behind Bracket Bots, a self-balancing robot for under $200 that can roam around your house.

Long-term memory with multimodal knowledge graphs to search 7 days of video and audio within 5 seconds. Winners @ Hack the North 2023.

Self-driving car design team, WATonomous, in Waterloo. I'm working on a low-latency data-driven controls model for the car.

Worked with Aviato to build a recommendations system for recruiting top AI talent amongst 300k+ engineers.

Worked with Tempo Labs to build generative UI agents in multiple programming languages concurrently.

Inpainting with diffusion language models to fill in missing text. In Progress.

Jailbroke a Galaxy S24 to run Moondream 3B VLM locally, with quantization + local linux setup on phone. Built at TreeHacks 2025

Trained a smaller VQGAN+CLIP to generate images from text and poetry prompts. Scaled up inference to run on MacBooks efficiently.

256 Dimension audio embeddings for semantic audio analysis using waveforms and Fourier Transforms, trained with contrastive learning.

Worked on safety simulation software for New York trains with custom network protocols.

Generating policy recommendations with AI agents from citizen complaints.