Soumya Sanyal
Applied Scientist at Amazon Web Services (AWS)
California, United States
Summary
👤
Senior🎓
Top SchoolSoumya Sanyal is an Applied Scientist at AWS building safety and redaction solutions for LLMs on the Amazon Bedrock Guardrails team, drawing directly on a Ph.D. in Computer Science from USC where his research advanced explainability, explanation-based learning, and robust reasoning in AI. With a decade of experience spanning research internships at Microsoft Research and multiple AWS terms, plus earlier production risk engineering at Goldman Sachs, he bridges rigorous research with product-grade safety systems. His work includes practical systems for hallucination detection, synthetic data generation, and improving logical and commonsense reasoning in large models. Comfortable across graph neural nets, NLP, and high-stakes finance infrastructure, he focuses on making agentic AI tools provably safer for real-world use. An often overlooked strength is his track record of shipping both research-driven prototypes and production risk engines, enabling rapid translation from novel algorithms to operational safeguards.
10 years of coding experience
2 years of employment as a software developer
Doctor of Philosophy - PhD, Computer Science, Doctor of Philosophy - PhD, Computer Science at University of Southern CaliforniaSenior Secondary, 91%, Senior Secondary, 91% at DAV Public SchoolBachelor of Technology (B.Tech.), Electrical, Electronics and Communications Engineering, Bachelor of Technology (B.Tech.), Electrical, Electronics and Communications Engineering at IIT Kharagpur
English, Hindi, Bengali, German
Github Skills (78)
graph-representation-learning10
nlp9
python9
deep-learning9
graph-convolutional-networks9
neural-machine-translation9
pytorch9
graph9
natural-language-processing9
nvidia9
transformer8
speech-recognition8
jekyll-theme8
machine-learning8
jekyll8
Programming languages (6)
Github contributions (5)
soumyasanyal/soumyasanyal.github.io
Nov 2019 - Aug 2024
Homepage
Contributions:6 PRs, 81 pushes, 5 branches in 4 years 10 months
malllabiisc/ProteinGCN
Apr 2020 - Oct 2020
ProteinGCN: Protein model quality assessment using Graph Convolutional Networks
Contributions:17 commits, 3 PRs, 15 pushes in 6 months
pytorchgraph-convolutional-networksgeometric-deep-learningdeep-learningconvolutional
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.