Sergei Shumilin
Research And Development Engineer at WANNA
Portugal
Summary
👤
Senior🎓
Top SchoolSergei Shumilin is an R&D engineer with 8+ years designing and shipping deep learning and computer vision systems for AR and virtual try-on experiences. He has led end-to-end projects from dataset and annotation pipelines to real-time rendering engines, notably delivering virtual makeup and footwear try-on features that combine NeRF/SDF reconstruction, Gaussian splatting, and geometry morphing for improved realism and occlusion handling. At WANNA he prototyped production-ready 3D asset reconstruction and integrated advanced splatting techniques into try-on flows, and previously built a real-time face reconstruction and iOS SDK for VOIR/Farfetch. Comfortable across research and product engineering, he refactors ML pipelines for scalability while keeping a pragmatic eye on real-time constraints. Based in Portugal with a background in applied mathematics, he often blends classical graphics priors with modern neural representations to push practical AR realism.
8 years of coding experience
10 years of employment as a software developer
Bachelor's degree, Applied Mathematics, Bachelor's degree, Applied Mathematics at Saint Petersburg State University
Github Skills (12)
geometry8
mesh8
python7
searching7
sparse5
latex4
pytorch3
tex3
sparse-matrix2
slurm1
supercomputer1
autograd1
Programming languages (3)
Github contributions (5)
SergeiShumilin/grid-to-grid-interpolation
Apr 2020 - Nov 2021
Interpolate data from one mesh to another using centers of cells for searching neigbhours
Contributions:66 commits, 42 pushes, 1 branch in 1 year 6 months
interpolatemeshcellssearching
SergeiShumilin/pyJSCC
May 2018 - Aug 2018
Sharing code within Joint SuperComputer Center
Contributions:7 PRs, 72 pushes, 10 branches in 3 months
supercomputer
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.