Aryan Barsainyan
Research Intern at Max Planck Institute for Biogeochemistry
Bengaluru, Karnataka, India
Summary
🤩
Rockstar🎓
Top SchoolAryan Barsainyan is a research-focused machine learning engineer who builds reproducible, uncertainty-aware frameworks for scientific domains spanning chemistry, biology, materials, environmental science, and olfaction. With ~4 years of hands-on experience across academia and industry—including contributions to DeepChem, Deep Forest Sciences, and a visiting scholar role at NUS—he bridges full-stack engineering, large-scale ML screening (1B+ datapoints), and principled research. He has authored peer-reviewed work in RSC Digital Discovery and ACS journals and helped develop benchmarks like STORI to evaluate learning in stochastic environments. An active open-source contributor, Aryan enhanced DeepChem’s D-MPNN featurizers and tooling to improve reproducibility and model robustness. He combines a mechanical engineering background from NITK with practical experience in cloud infrastructure, CI/CD, and mentor roles (GSoC), aiming to make scientific AI verifiable and broadly generalizable.
4 years of coding experience
4 years of employment as a software developer
Jindal Vidya Mandir, Vidyanagar, JSW TownshipBachelor of Technology - BTech Mechanical Engineering, Bachelor of Technology - BTech Mechanical Engineering at National Institute of Technology Karnataka
Github Skills (10)
pytorch10
machine-learning10
deep-learning10
graph-neural-network10
python10
data-science10
biology9
drug-discovery9
rdkit9
unit-testing8
Programming languages (2)
Github contributions (5)
deepchem/deepchem
Jun 2022 - Dec 2022
Democratizing Deep-Learning for Drug Discovery, Quantum Chemistry, Materials Science and Biology
Role in this project:
Data Scientist
Data ScientistContributions:117 reviews, 54 commits, 51 PRs in 7 months
Contributions summary:Aryan's commits primarily focus on resolving deprecation warnings, improving code formatting, and adding new functionality related to atom features and bond features for the DMPNN featurizer. They also added a global feature generator and implemented a mapper class, further expanding the capabilities of the DMPNN model. Additionally, the user worked on fixing bugs related to graph data and unit tests.
deep-learningquantum-chemistrydrug-discoverybiologymaterials-science
ARY2260/deepchem
Jun 2022 - Jan 2026
Democratizing Deep-Learning for Drug Discovery, Quantum Chemistry, Materials Science and Biology
Contributions:4 PRs, 159 pushes, 38 branches in 3 years 7 months
deep-learningquantum-chemistry
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.