- 🎓 B.Tech in Computer Science and Engineering, IIITDM Jabalpur — expected May 2027
- 🤖 Primary focus: Agentic AI — multi-agent LLM pipelines, confidence-gated reasoning loops, and self-correcting reflection agents
- 💼 Selected for a competitive Future Founders internship at OkCredit, where I shipped a native Android credit-risk product end to end
- 🧠 Solved 300+ DSA problems on LeetCode and other competitive platforms
- 📍 Based in Jabalpur, Madhya Pradesh, India
- 🎯 Currently seeking Software Engineer / SDE / Full-Stack Developer internships
- 💬 Ask me about LLM agent architectures, Supabase realtime, or offline-first Android
| Project | What it does | Stack | Link |
|---|---|---|---|
| CivicLoop | A single citizen photo triggers an autonomous multi-agent LLM pipeline — categorization, deduplication, routing, auto-escalation — on a confidence-gated Sense → Plan → Act → Reflect loop, with a Reflection agent that re-routes its own misclassifications. Every decision is published to a realtime, append-only public Accountability Ledger. | React 19 · TypeScript · Supabase · PostgreSQL · Google Gemini · Firebase | Repo |
| Vasooli — Collection Radar | Native Android app that scores retailer credit risk 0–100 from a six-signal weighted engine computed entirely from a wholesaler's own ledger. An expected-loss model (risk × outstanding exposure) turns a static balance book into a daily ranked collection list. Fully offline-first, every score explainable in plain language. | Kotlin · Jetpack Compose · Room · SQLite | Repo |
Engineering highlights
- 🛰️ Serverless SLA-Watcher Firebase Cloud Function that auto-escalates any civic report once its response deadline lapses
- 🔐 Row-Level Security + Realtime subscriptions on Supabase for a tamper-evident public ledger
- 🗺️ Google Maps and Street View integrated into the React 19 frontend over REST APIs and realtime streams
- 📊 Custom Canvas-rendered visualizations in Android — aging chart, sparklines, risk gauge — with zero external charting libraries
OkCredit — Future Founders Intern (Software Development) · Remote · May 2026 – Jul 2026
Selected for an 8-week, fully remote product-and-engineering internship across 3 phases and 8 deliverables. Conducted structured user research with 10 retail merchants to profile their ledger and collection workflows, identified collection prioritization as the highest-impact problem, then owned requirements, architecture, implementation, and testing for Vasooli — the internship's primary engineering deliverable.
Languages
AI & LLM
Frontend
Backend
Databases
Tools
- 🧮 300+ data structures and algorithms problems solved on LeetCode and other competitive platforms
- 📈 All India Rank 14,350 in JEE Main
- 🎤 Campus Ambassador, Buyhatke (NextGen Program) — drove student engagement and mobile app sign-ups through on-campus outreach
- 📜 Programming for Everybody (Getting Started with Python) — University of Michigan on Coursera, Dec 2024
- 🎓 95% in Class XII, Bhashyam Junior College
💬 Always up for a conversation about agentic AI, systems that explain their own decisions, or anything worth building.