NASA International Space Apps Challenge 2025 – Best Use of Data (Global Winner)
Team “Resonant Exoplanets” – AI system for exoplanet and biosignature detection from telescope and satellite spectra.
View project →Awards
Projects
Resonant Worlds Explorer
NASA Space Apps Global Hackathon Winner · Oct 2025
AI-powered exoplanet detection and biosignature scoring system built on NASA datasets using ML pipelines and physics-based validation.
- Largest Global Hackathon Winner (Python, React, FastAPI, Cloud Run).
- Built an AI-powered system for exoplanet detection and biosignature scoring on NASA datasets using ML pipelines and physics-based validation.
- Developed an end-to-end platform with a FastAPI backend and React frontend, deployed on Cloud Run with logging and automated data refresh to support interactive exploration of model outputs.
- Python
- React
- FastAPI
- Cloud Run
Research & Technical Reports
Practical Lessons from Building ML-Powered Code Intelligence and Reliability Workflows for an Early-Stage Developer Tools Platform
Independent technical note · 2026
Sahil San
Independent technical report on hybrid code intelligence, semantic retrieval, and AI-assisted reliability workflows for PRs, CI logs, issues, and static analysis.
Read PDF →Experience
Founding Software Engineer
Stealth Startup
Jun 2025
New York, NY
- Built ML workflows to detect and count electrical devices from construction-plan PDFs, supporting automated takeoff review for contractors and lighting vendors.
- Developed validation workflows that help estimators review AI-generated device counts before export.
Machine Learning Engineer Intern
IBM
May 2026 – Sep 2026
New York, NY
- Built ML-driven synthetic transaction generators on Red Hat OpenShift to validate hybrid-cloud integrations, reducing workload runtime from ~10 minutes to ~30 seconds.
- Integrated automated validation workloads into CI pipelines with QA and SWE teams, improving end-to-end test coverage and release confidence for hybrid-cloud deployments.
Software / ML Engineer
StarDrive
Mar 2026 – Aug 2026
San Francisco, CA
- Developed simulation, control, and motion-planning components for FieldSpace, a deterministic physics-based autonomy platform for warehouse and forklift workflows.
- Built and tested services for experimental physics-based AI compute systems, supporting model integration, deployment, and evaluation in early R&D environments.
Founding Machine Learning Engineer
DevPlaza
Jan 2026 – Aug 2026
San Francisco, CA
- Built end-to-end ML pipelines for code intelligence, search, and recommendations using Python, including model serving, monitoring, and production-feedback iteration.
- Improved developer-facing code retrieval through semantic search and reliability-focused workflows for an early-stage developer tools platform.
Machine Learning Researcher
Waymo
May 2025 – Aug 2025
Mountain View, CA
- Designed synthetic-data-generation pipelines to train and evaluate ML models, improving coverage of edge cases beyond limited real-world data.
Skills
Languages
- Python
- JavaScript
- TypeScript
- SQL
- Bash
ML / AI
- PyTorch
- TensorFlow
- Model Evaluation
- Synthetic Data
- Model Serving
Systems
- Docker
- Kubernetes
- OpenShift
- CI/CD
- Microservices
- FastAPI
- Linux
- Git
- GitHub
Education
University of California San Diego
B.S., Artificial Intelligence
Relevant Coursework
- Data Structures & Algorithms
- Discrete Math
- Linear Algebra
- Object Oriented Programming
- Machine Learning
- Natural Language Processing
- Probability & Statistics