Matthew Tancik
Head Of Applied Research at Luma AI
Menlo Park, California, United States
Summary
🤩
Rockstar🎓
Top SchoolMatthew Tancik is Head of Applied Research at Luma AI with 11 years of experience bridging cutting-edge computer vision research and production engineering. He holds advanced degrees from MIT and is pursuing a PhD at UC Berkeley, combining strong theoretical training in EECS and physics with hands-on machine learning systems work. His open-source contributions include foundational components for nerfstudio—a prominent NeRF toolkit—where he implemented encodings, camera models, samplers, renderers, and occupancy-grid support, and experimental work on Fourier feature networks for high-frequency function learning. Earlier internships at Waymo, Cruise, and Google reflect deep experience in perception and large-scale engineering environments. Colleagues know him for shipping core model infrastructure that enables rapid experimentation and for translating research ideas into robust backend systems. He thrives at the intersection of research and applied productization, often favoring elegant, reusable abstractions that accelerate teams.
11 years of coding experience
Doctor of Philosophy - PhD, Doctor of Philosophy - PhD at University of California, BerkeleyMaster of Engineering - MEng, EECS, Master of Engineering - MEng, EECS at Massachusetts Institute of Technology
Github Skills (11)
neural-network10
computer-vision10
machine-learning10
jupyter-notebook10
pytorch10
serf10
deep-learning10
python10
3d-reconstruction10
google-colaboratory9
linear-algebra9
Programming languages (5)
Github contributions (5)
nerfstudio-project/nerfstudio
Mar 2022 - Jan 2023
A collaboration friendly studio for NeRFs
Role in this project:
Back-end Developer & ML Engineer
Back-end Developer & ML EngineerContributions:26 releases, 843 reviews, 516 commits in 10 months
Contributions summary:Matthew's contributions are focused on building the foundational components for a NeRF-based studio. They implemented the core structure for several modules, including encodings, camera models, samplers, and renderers, demonstrating an understanding of computer vision concepts. Additionally, the user added MLP model, render heads, and created a base class for models as well as initial support for an occupancy grid.
pytorchphotogrammetrydeep-learningcomputer-vision3d-reconstruction
tancik/fourier-feature-networks
Jun 2020 - Jan 2023
Fourier Features Let Networks Learn High Frequency Functions in Low Dimensional Domains
Role in this project:ML Engineer
ML EngineerContributions:22 commits, 2 PRs, 5 pushes in 2 years 7 months
Contributions summary:Matthew's commits primarily involve modifications to Jupyter notebooks focused on machine learning experiments. These notebooks explore various techniques, including 2D image regression, 1D regression, and 3D MRI, within the context of Fourier Feature Networks. The commits indicate an active role in developing and testing different models and configurations. The presence of colab links in commit messages suggest active work using Google Colaboratory.
dimensionalhigh-frequencyfourierletdomains
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.