RelayOS
AI decision memory platform. A RAG-based conflict detection system flags contradicting decisions (pgvector search to shortlist, LLM judge to confirm), and a hand-written MCP server lets AI clients query stored decisions directly.
Full-Stack & AI Developer
I build and ship production AI systems independently — from RAG pipelines and MCP servers to CI-integrated developer tools.
aadi198555@gmail.comEducation
Manipal University Jaipur
Master of Computer Applications (MCA)
Jul 2026 — Present
University of Rajasthan
Bachelor of Computer Applications (BCA)
Aug 2022 — Jul 2025
Indian Institute of Technology, Mandi
Professional Certification
Apr 2025 — Jun 2026
About
Full-stack and AI engineer who ships and maintains production systems independently. Built RelayOS, an AI decision memory platform with a hand-written MCP server and RAG-based conflict detection, and Preflight, a dependency-upgrade risk scanner using AST parsing with a CI/CD integration. Works across Next.js, TypeScript, FastAPI, PostgreSQL, and LLM integrations. Maintains 50+ GitHub repositories with merged PRs on production codebases.
Projects
AI decision memory platform. A RAG-based conflict detection system flags contradicting decisions (pgvector search to shortlist, LLM judge to confirm), and a hand-written MCP server lets AI clients query stored decisions directly.
Dependency upgrade risk scanner. A CLI that uses Babel AST parsing to flag exactly which files will break before an upgrade, plus a GitHub Actions integration that comments risk on PRs and fails CI on breaking changes. 87 tests, published to npm.
AI-powered real-time communication platform with transcription and channel insights using the OpenAI API; sub-250ms latency, 5–10 concurrent users.
RAG pipeline with OpenRouter integration routing queries across 2–3 LLM models per request for improved relevance.
Experience
Aug 2026 — Oct 2026
Built and debugged core flows (trip planning, itineraries, saved trips, sharing, auth) for an AI-powered travel platform. Integrated the Calendly API and built backend services and REST APIs, resolving production bugs across the stack.
Apr 2026 — Jun 2026
Built end-to-end full-stack features across frontend, backend, and AI-integrated systems in a fast-paced startup, integrating AI APIs into production-ready apps with modern JavaScript and TypeScript frameworks.
Jan 2026 — Mar 2026
Merged 8+ pull requests across production codebases, resolving backend issues, integrating webhook-based features, and standardizing error handling across 5+ API endpoints. Architected centralized middleware handling 100% of API responses.
Skills