Sameer Chaturvedi
ML + software
engineer.
Machine Learning Engineer at Amazon Prime Video and Lead Software Engineer at MIT's Plasma Science and Fusion Center. Before that, four years at Microsoft on petabyte-scale security ML.
I studied computer science and astrophysics, and I've spent my career where ML meets something I care about: recommendation systems, fusion research, cyber security, and computer vision for basketball.
Currently
Amazon Prime Video
Machine Learning Engineer II
Prime Video's personalization service assembles the recommendation rows on the home page and storefronts worldwide. I work across the ML models and data pipelines that rank and mix those rows, and I'm a core developer of the agentic AI tooling the team uses to operate them.
MIT Plasma Science and Fusion Center
Lead Software Engineer
DisruptionPy is the group's open-source framework for turning raw tokamak experiment data into analysis-ready and ML-ready datasets across five machines. I own its software architecture: the data-access layer, configuration, multiprocessing, and developer tooling.
Work
From container runtimes
to recommendation models.
Production ML at Amazon, scientific software at MIT, four years of security engineering at Microsoft, and earlier stints in live basketball analytics, on-device computer vision, and Kubernetes internals.
Amazon
Recommendation ML + agentic AI
Machine Learning Engineer II / Prime Video Personalization and Discovery, New York
Prime Video's personalization service assembles the recommendation rows on the home page and storefronts worldwide. I work across the ML models and data pipelines that rank and mix those rows, and I'm a core developer of the agentic AI tooling the team uses to operate them.
Read moreMIT PSFC
Scientific software for fusion
Lead Software Engineer / Disruption Studies Group -- DisruptionPy
DisruptionPy is the group's open-source framework for turning raw tokamak experiment data into analysis-ready and ML-ready datasets across five machines. I own its software architecture: the data-access layer, configuration, multiprocessing, and developer tooling.
Read moreCerebro Sports
Short contract, live stats demo
Machine Learning Engineer, contract / Live Gameplay Team
A brief contract around the NBA Summer League in Las Vegas. I built a demo app for exploring live basketball tournament stats conversationally: ask a question in plain English, get the numbers, realtime graphs, and dashboards back.
Read moreMicrosoft
Security ML at petabyte scale
Software Engineer II / M365 Security Detections
Four years on the petabyte-scale platform that detects cyber-attacks across all of M365 in near real time. Lead engineer for Microsoft's first mail anomaly detection and for the GPT-based detection assistant.
Read moreAI.Reverie
On-device computer vision
Machine Learning Engineer / Computer Vision Team
Real-time object detection and instance segmentation on iOS: PyTorch models converted to LibTorch and run on-device through a Swift + AVFoundation capture pipeline.
Read moreRed Hat
Kubernetes container runtimes
Software Engineer / OpenShift -- Container Runtimes
Upstream contributions to CRI-O and Libpod, the container runtime and pod manager underneath OpenShift and Kubernetes.
Read moreSide projects
Outside
of work.
Muse is the long-running one: computer vision for basketball coaching, scouting, and training. The name comes from basketball being my muse for upskilling technologically and practicing discipline.
