I build intelligent systems for complex, real-world domains.
AI/ML undergraduate at SRM Chennai focused on systems engineering, real-time software, structured reasoning, and reliable AI.
I like building systems where data, dependencies, and decisions are explicit, testable, and measurable.
60 Hz F1 telemetry → ingestion → persistence → real-time streaming → performance analysis → AI debriefing
- Stack: Python, FastAPI, PostgreSQL, TimescaleDB, Next.js, Socket.IO, FastF1, UDP
- Supports F1 2020–2025 telemetry through modular packet adapters
- High-frequency telemetry ingestion and persistence
- Live driver/performance analysis across braking, corners, lap deltas, thermal/ERS behavior and battles
- FastF1 real-world reference data integration
- Automated quality gates with unit + PostgreSQL integration tests, linting, type checking and production builds
Graph-based diagnosis and bounded recovery for payment failures
- Stack: Python, FastAPI, React/TypeScript
- Models payment dependencies as an explicit graph
- Traces failures to root causes using per-edge evidence
- Selects bounded recovery actions and measures outcomes
- Deterministic simulator with reproducible evaluation
- Uses held-out validation, fair baselines and counterfactual analysis
Blockchain-based ticketing system focused on verifiable ownership and controlled redemption.
- Stack: TypeScript, Solidity, Polygon, Next.js, AWS, Docker
- Cryptographic ticket verification
- Expiring QR codes and one-time redemption
- Smart-contract backed ticket lifecycle
- Analytics dashboard and cloud deployment
Languages: Python, C++, TypeScript, Java, JavaScript, SQL
Backend: FastAPI, Node.js, Express, Django
Frontend: React, Next.js
Systems & Data: PostgreSQL, TimescaleDB, UDP, real-time streaming, data pipelines
Cloud & DevOps: AWS Lambda, S3, API Gateway, CloudWatch, Docker, CI/CD
Engineering: Systems design, distributed systems, testing, observability, performance optimization, graph algorithms
- Structures over magic - model systems explicitly and make reasoning inspectable.
- Determinism when it matters - reproducible inputs should produce reproducible decisions.
- Evidence over assertions - every important decision should have a traceable basis.
- Fair evaluation - strong baselines, held-out validation and transparent methodology.
- Design discipline - document decisions, test assumptions and make failures visible.
Infinitra Innovations - Full-stack Development Intern
Building cloud applications, data infrastructure and production AI systems with AWS.
AWS Student Builder Group @ SRMIST - AI & ML Associate
Leading a student engineering team and driving technical execution across projects.