- 🎓 B.Tech at IIT Roorkee (2023–2027)
- 💻 Focus areas: Software Engineering with applied AI — agentic AI systems (Langchain & LangGraph)
- 💼 Software Development Intern @ Jivika — optimized features on companies internal admin portal (Next.js, Node.js) for an AI-native hiring platform
- 🌱 Open-source contributor @ FlowiseAI (YC S23, 53k★) — 3 merged PRs
- 🏆 Specialist on Codeforces
- 🏆 Top 6% on LeetCode Globally
- 🏆 3★ on CodeChef
A GitHub-style platform built from scratch: custom CLI for version control, AI-powered duplicate-issue detection, and real-time collaboration.
- Architected a full-stack VCS with 35 REST endpoints across 4 MongoDB models — repos, issues, commits, notifications
- Built a custom Git-like CLI (
init/add/commit/push/pull/revert) with local staging, pushing payloads up to 50MB to AWS S3 - Designed a Gemini + Pinecone RAG pipeline over issue text to flag semantic duplicates, cutting manual triage by 40%
- Built a Socket.IO notification layer with private per-user rooms — instant cross-session alerts, zero client-side polling
- Added WebRTC peer review: any commit can be reviewed live over a shared video call
- JWT-secured REST API; deployed on AWS Amplify (frontend) + EC2 (backend)
📂 Repo · 🎥 Demo Video · 🔗 Live App
A production-grade, server-rendered catalog admin panel with role-based access, AI content generation, and live analytics.
- Built a server-side rendered dashboard in Next.js 16 (App Router + Server Components) with full CRUD for 25+ SKUs, cutting initial load time by 40%
- Implemented multi-provider auth (Email, GitHub, Google) via NextAuth.js v5 with 3-tier RBAC enforced through middleware route protection
- Integrated AWS S3 presigned URLs for direct browser-to-S3 image uploads, removing all server bandwidth overhead
- Built an AI description generator on OpenRouter with a 3-model fallback chain (LLaMA 3.2, Gemini 2.0 Flash, Phi-3 Mini)
- Shipped a real-time analytics suite (MongoDB aggregations → 4 Recharts dashboards) for 90-day sales and low-stock trends
- Dockerized with multi-stage builds (75% smaller image); SSR metadata tuned for AI-crawler and social-share SEO
Languages
Frontend
Backend & Databases
AI / ML, Agents & Vector Search
Cloud & DevOps
Open to full-time Applied AI Software Engineering / Applied AI Full-Stack Engineering roles — remote or in-office. Reach out: sahil.kamble.work@gmail.com · Portfolio