We're a startup building computer vision video analysis systems!Backed by Pear VC

Joined the recently acquired OrdinalHub team that specializes in tracking, analyzing, and aggregating Inscriptions:- Developed payment/invoice system using BtcPay and Typescript (tRPC) for the Bitcoin Ordinal Inscription Service. - Sped up and implemented new data aggregation operations for the Inscription Indexer Client, improving API latency for 1000+ enterprises by 800%.- Developed Quadkey project (with a gateway through the Developer API) focused on ordinal enveloping, allowing for improved alignment and consensus on new metaprotocols, enhancing consistency and speeding up market delivery while protecting users.

Designed and developed a profitable cross-chain arbitrage MEV Bot in Typescript (prototype) and Golang (MVP) that generates additional passive revenue from Luxor’s mining pool infrastructure.- Reverse-engineered the decentralized exchange protocols UniswapV2 & UniSwapV3 and developed a library to stream data and accurately assess the impact of incoming mempool transactions on market makers.- Worked directly with industry researchers to design a proprietary method that optimally extracts profit from incoming block templates.- Implemented a system that processes metadata of all mempool transactions and exchanges data from CEXs/DEXs, finds optimal arbitrage swap path with a custom Dijkstra variant, and executes profitable cross-chain opportunities with <100ms latency. Proven to be able to generate $50K+ profit in a day.- Made sure the system was reliable and observable in production with Kubernetes, Prometheus, and Grafana stack

Developed computer vision (CV) software on a team with AWS software engineers for a smart dashcam stealth startup that provides enhanced car safety ratings for auto insurance companies by automatically detecting and reporting infractions. - Created proprietary methods (using traditional CV and low-cost deep learning) for detecting cars, lanes, and license plates on NVIDIA Jetson hardware utilizing OpenCV & Tensorflow in real-time at 60+ FPS.- Built an online multi-object tracking (MOT) system for cars with a metric learning-based re-identification system utilizing appearance and contextual clues.

Worked with a small, fast-paced startup team building a social media platform for 360 virtual tours.- Developed the frontend (React) and backend (PHP) for the login/signup system, real-time chat system with WebSockets, and various app features (such as follow/like/share).- Used Three.js to efficiently render Equirectangular images and allow mouse/keyboard controlled movement- Designed the entire data schema for the MySQL database