Hayk Martiros
Research at Google DeepMind
San Francisco, California, United States
Summary
🤩
Rockstar🎓
Top SchoolHayk Martiros is a research-focused engineering leader with 13 years building autonomy, robotics, and creative AI systems across startups and top labs. He co-founded Producer.ai to reinvent music creation and now co-leads DeepMind’s Lyria project after the team's acquisition, blending product vision with deep research. Previously as VP of Autonomy at Skydio he scaled visual-autonomy stacks for real-world inspection and security, moving from hands-on autonomy engineer to executive leadership. Hayk contributes open-source work that bridges ML and audio—most notably backend and frontend work on Riffusion’s real-time music generation and spectrogram-to-waveform pipelines. He holds degrees from Princeton and Stanford in engineering and robotics, and is known for turning complex control and perception research into deployable systems. Peers describe him as a builder who prefers tackling hard engineering problems collaboratively and shipping delightful, unconventional products.
13 years of coding experience
5 years of employment as a software developer
Master of Science, Robotics, Master of Science, Robotics at Stanford UniversityBS in Engineering, ME / Robotics / CS, BS in Engineering, ME / Robotics / CS at Princeton University
Stackoverflow
2,176reputation
482kreached
30answers
9questions
css
top-5%
Github Skills (30)
threejs10
pytorch10
python10
typescript10
audio-processing10
stable-diffusion10
diffusers10
react-three-fiber10
audio10
tailwindcss10
ai10
typescript-types10
typescripts10
music10
react10
Programming languages (8)
Github contributions (5)
riffusion/riffusion-hobby
Nov 2022 - Feb 2023
Stable diffusion for real-time music generation
Role in this project:
Back-end Developer & ML Engineer
Back-end Developer & ML EngineerContributions:1 release, 9 reviews, 127 commits in 2 months
Contributions summary:Hayk primarily contributed to the development of the Riffusion pipeline, focusing on audio generation and image-to-audio conversion. They implemented data model specifications for the API and integrated image-to-image aware interpolation. Further contributions include building modules for spectrogram and waveform processing, and setting up a Flask server for API serving. They also added support for masks, and integrated SSL and caching features.
pytorchmusic-generationmachine-learningaudioai
riffusion/riffusion-app-hobby
Nov 2022 - Dec 2022
Stable diffusion for real-time music generation (web app)
Role in this project:Front-end Developer
Front-end DeveloperContributions:2 reviews, 114 commits, 9 PRs in 1 month
Contributions summary:Hayk primarily focused on building the user interface and front-end components of the real-time music generation web app. Their commits demonstrate the setup and configuration of React, Three.js, TypeScript, and Tailwind CSS for UI development. The user implemented key UI elements, including interactive 3D visualizations using React Three Fiber, and integrated the display of spectrograms. They also contributed to setting up the basic layout and adding interactive elements like buttons.
reactmusic-generationaudiothreejsdiffusion
Find and Hire Top DevelopersWe’ve analyzed the programming source code of over 60 million software developers on GitHub and scored them by 50,000 skills. Sign-up on Prog,AI to search for software developers.