Soran Ghaderi
Machine Learning Research Engineer at CoSTAR Network
London, England, United Kingdom
Summary
🤩
Rockstar🎓
Top SchoolSoran Ghaderi is a Machine Learning Research Engineer based in the UK with eight years of experience at the intersection of generative video, energy-based models, and scalable ML systems. Currently at CoSTAR Network via Abertay University, he combines academic rigor from an MSc in AI with hands-on open-source engineering—authoring CUDA-accelerated randomized linear algebra, transformer utilities, and libraries for energy-based and graph models. His research trajectory spans practical code-generation guides for LLMs (NIR) and diffusion/flow model scaling at UIUC, reflecting a rare blend of theory, systems optimization, and experiment-driven evaluation. Comfortable moving between C++/CUDA performance work and Python research stacks, he actively seeks PhD opportunities and collaborations that bridge mathematics, neuroscience, and machine learning.
8 years of coding experience
2 years of employment as a software developer
Master of Science - MS, Artificial Intelligence, Master of Science - MS, Artificial Intelligence at University of EssexBachelor of Engineering - BE, Computer Software Engineering, Bachelor of Engineering - BE, Computer Software Engineering at University of Kurdistan, Sanandaj
English, Kurdish
Stackoverflow
237reputation
717kreached
3answers
0questions
Github Skills (89)
transformers10
graph-algorithms10
python10
attention-mechanism10
diffusion-models10
deep-learning10
self-attention10
gemma10
multihead-attention10
glm10
multimodal10
audio10
neural-network10
transformer10
nlp10
Programming languages (4)
Github contributions (5)
soran-ghaderi/EfficientCoF
Jan 2018 - Jan 2020
A new efficient subspace and K-Means clustering based method to improve Collaborative Filtering
Contributions:89 commits, 2 PRs, 58 pushes in 2 years
collaborative-filteringk-means-clusteringrecommender-systemsparsityhigh-dimensionality
soran-ghaderi/torchebm
Oct 2024 - Jul 2026
🍓 Simulation-free, GPU-first generative modeling in PyTorch ⚡ Composable primitives for scalable, stable training of modern EBMs, diffusion, flow matching, and Schrödinger bridges.
Contributions:15 reviews, 96 PRs, 112 pushes in 1 year 9 months
gpupytorchsimulationcontrastive-divergencecuda
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.