name: Aditya Singh
location: Pune, India
focus:
- RAG pipelines, LLMs and agent workflows
- Backend and distributed systems
- Full-stack web apps
previously:
- Software Engineer @ Stock Unlock (YC W22), 2025-2026
- Associate Software Engineer @ CDK Global, 2024-2025
- Software Engineer Intern @ CDK Global, 2023-2024
looking_for: Backend, AI engineering or full-stack roles. Remote or on-site.| Project | What it does | Stack |
|---|---|---|
| docsense | Document Q&A backend: hybrid pgvector and full-text retrieval fused with RRF, reranking, cited answers, semantic cache, RAGAS evals | Python, FastAPI, pgvector, Celery, Langfuse |
| cua-computer-use-automation | An LLM agent explores a web app once, then the run is compiled into a script that replays without a model, with human takeover | Python, Playwright, Pydantic, FastAPI |
| delta-chat | Diffs two revisions of an engineering drawing and answers questions about the changes with citations, or declines | Python, OpenCV, Tesseract, BM25, OpenTelemetry |
| synthsense | LangGraph agents triage, diagnose and suggest fixes for incidents, streamed live to the browser over SSE | Python, LangGraph, FastAPI, pgvector, React |
| rtsp-viewer | Watch RTSP camera streams in the browser: one FFmpeg process per camera, fanned out to every viewer over WebSockets. Live demo | Go, FFmpeg, React, TypeScript |
| medical-report-ocr-service | Turns lab report PDFs into FHIR R4 Observations with LOINC and UCUM codes. Deployed instance | Node.js, TypeScript, Express, zod |
Languages and frameworks
Data and infra
AI