Full-Stack Engineer · MERN Stack · AI Systems · Distributed Systems
Third-year CS @ IIIT Nagpur (CGPA: 8.12) · Upcoming SWE Intern @ SMS DataTech Japan
I build and ship real products — AI platforms, encrypted real-time systems, and distributed backends.
Upcoming SWE Intern at SMS DataTech Japan. Previously at QuickIntell building agent-based systems with AWS SQS, Telnyx, and Deepgram. 600+ LeetCode problems. Two national hackathon placements.
Exploring: System Design · RAG pipelines · MCP · OpenCV · Custom Encryption
Sept 2025 – Present
- Migrated state management from React Context to Zustand → 35% fewer redundant API calls
- Architected agent-based backend with AWS SQS for concurrent document uploads and async processing across workers
- Integrated Telnyx APIs for mass outbound calling campaigns serving 100K+ users per campaign
- Built cursor-based pagination with Prisma → 40% faster page loads
- Built webhook-driven call analytics pipeline using Inngest + S3-triggered events → 40% faster setup and deployment
- Developed call analytics platform leveraging Deepgram STT for background processing of recorded calls
Jan 2025 – Mar 2025
- Enhanced PHP/Hack MVC backend handling 1,000+ daily requests
- Optimized routing logic → 15% faster average response times
- Refactored data flow pipelines → 20% higher throughput and reliability
- Improved codebase structure and documentation → 25% less maintenance overhead
React.js Node.js MongoDB React Native Selenium BeautifulSoup Tailwind CSS
AI-powered outfit recommendations based on weather, location, age, and gender — with a custom 3D model try-on feature. Scrapes 1,000+ products from Amazon and Myntra.
| Metric | Result |
|---|---|
| Outfits generated in testing | 300+ |
| Shared looks in month one | 120+ |
| Marketplace listings (month 1) | 40+ |
| Daily active users (post Android launch) | 100+ |
| Uptime | 99% |
Custom load balancer + AWS SQS pipeline for distributed scraping across multiple servers.
Node.js Express.js React.js MongoDB Socket.IO Docker P2P AES-256-CBC SHA-256
P2P distributed file storage system with sharding, end-to-end encryption, and fault-tolerant recovery.
| Metric | Result |
|---|---|
| Max file size supported | 1GB+ |
| Throughput improvement (heartbeat peer mgmt) | ~2x |
| Peer dropout reduction | 40% |
| Media size reduction (WebP pipeline) | 90% |
| Storage cost reduction | 60% |
| Client load time improvement | 35% |
| Data recovery from partial shard loss | 100% (up to 30% loss) |
100KB sharding with SHA-256 integrity checks, Reed-Solomon erasure coding, and AES-256-CBC encryption end-to-end.
| 🥇 5th Place | National Hackathon · IIITM Gwalior (2024) — AI precision farming platform |
| 🏅 Top 50 / 1,600 | Medecro.ai National Hackathon (2024) |
| 🎯 Semi-finalist | GRAB Hackathon (2025) |
| 📐 Qualified | IYMC (2023 & 2024) |
| 💻 600+ solved | LeetCode — strong DSA foundation |
Languages
Frontend
Backend
Databases & DevOps
AI & Cloud
Open to internships, collaborations, and interesting problems.