Software Engineer with iOS, Full-Stack, and AI Data Engineering experience.
4 years of experience in fast-paced startups: rapidly delivering ambitious projects, quickly adapting to new technologies, and collaborating in dynamic remote environments.
Passionate about building magical products and tackling complex problems, no matter the language or stack.
Seeking a collaborative, mission-driven remote team where I can make a real impact.
June 2021 - November 2024 — Multiple roles and products
Full-Stack Software Engineer
Hotshot — Make personalized generative AI images & videos
Part of a team of 4 that trained a Sora-level text-to-video model from the ground up in 4 months, generating HD video up to 10s at 24 fps: hotshot.co/release
Built and managed distributed data processing pipeline in Python to prepare large image and video datasets for model training, handling 1B+ samples and multiple petabytes of data.
Led end-to-end frontend and backend product development for iOS app and website, including subscription paywalls, user-friendly and moderated model inference, and a personalized invite system.
Led all iOS development of a TikTok-style app from the ground up, managing the iOS team, conducting code reviews, and architecting the app to align with product specifications.
Drove rapid product iteration using feature flags, A/B testing, and dynamic, server-controlled UI, enabling feature adjustments without requiring new client builds.
Individually contributed to full-stack feature development, building a custom framework for on-device non-custodial crypto wallets, and the “Collabs” feature (FaceTime + Avatars) with live audio and multiplayer 3D avatar interaction.
Quickly onboarded onto an extensive, pre-existing codebase, shipping daily App Store updates to millions of users while maintaining consistent stability and performance.
Led the full-stack development of “Parties”, a Clubhouse-style livestreaming feature, collaborating closely with a designer to refine the user experience and interface.
Developed a high-performance, real-time hand tracking system, leveraging a customized framework to process raw camera data into precise 3D coordinates.
Developed an iMessage-style app to enable anyone to easily chat with AI in a familiar interface, following Apple’s HIG for an intuitive and seamless user experience.
Integrated llama.cpp with a Swift layer to run models optimized for on-device use like Llama, Qwen, and Ministral, and proxied requests to remote LLM APIs.
Implemented in-app subscriptions with RevenueCat, automated build distribution and metadata management with Fastlane, and automated testing with XCTest and XCUITest.