BS Data Science student, IIT Madras · BCA graduate, 2025
I am interested in graph computation, machine learning, and the software behind reproducible research. My work connects mathematical ideas to implementations that can be tested, examined, and built upon.
Research · Open source · Selected projects · Academic direction
Sparse graph computation · Research software and experimental study
K GraphFuse explores reusable graph-filter plans, directional sensitivities, and optional GPU execution. The accompanying manuscript examines similarity-aware evaluation and selective GPU compensation. The project brings together the implementation, experimental results, and reproducibility materials.
Repository · Manuscript · PDF · Benchmark evidence
Technical preprint draft; not peer-reviewed. Results should be interpreted within the project's documented experimental scope.
View a figure from the controlled experiments
Controlled synthetic workloads on two GPUs; resident compute only. Follow the figure for methodology and limitations.
My contributions to shared infrastructure include the following merged work:
- OpenTelemetry: Browser Fetch client conformance.
- qtop: Scheduler error handling and terminal-size fallback and OAR file handling.
Additional OpenTelemetry work in review or development
As of November 2026: XMLHttpRequest conformance · MCP span classification · Entity identity coverage · Python GenAI test requirements · State-change conventions · LangGraph instrumentation (draft; dependent on the conventions proposal)
ReproLearn — Research reproducibility with inspectable, claim-level evidence.
AgentAgri — Agricultural information spanning crop health, weather, markets, and farmer-owned memory.
AXIOM Urban — Traffic-corridor analysis using FastAPI and a gRPC simulation engine.
SNAPV — Local-first screenshot exports across Chrome, Firefox, and Edge.
NeuroGolf 2026 — Compact ONNX solvers and reproducible visual-transformation experiments.
SciBoat (private repository) — Rust-backed continuity, reviewed execution, and receiver-attested replay.
I am seeking funded graduate research and scholarship opportunities in machine learning, data science, and research software. I hope to contribute to work grounded in careful evaluation, open collaboration, and useful, reproducible tools.
Selected tools: Python · TypeScript · Rust · React · FastAPI · Docker · PostgreSQL · ONNX · GitHub Actions