Backend AI Engineer in the making · B.Tech CSE @ Sister Nivedita University .
Building real systems at the intersection of AI, data, and infrastructure
Suvidha AI — AI platform empowering 15M+ gig workers in India with earnings forecasting (LSTM), demand heatmaps, bias detection (Isolation Forest), and a Hindi-first AI assistant via AWS Bedrock. Built for the AI for Bharat Hackathon.
ET Sentinel — Multi-signal convergence engine cross-correlating fundamental, technical, and sentiment signals across Nifty 200 every 2 hours. Shortlisted from ~55,000 teams at ET Gen AI Hackathon. Stack: Next.js 14, Python Lambda, DynamoDB, Angel One SmartAPI, GDELT + VADER.
ITMS for Indian Railways — Integrated Track Monitoring System built at Smart India Hackathon 2025 (Finalist). Real-time IoT-based track condition monitoring with predictive analytics and full-stack dashboard.
EXPO PAY — Cross-border payment platform on the Stellar blockchain with Universal IDs, smart escrow contracts, and fiat-to-crypto conversion.
Actively exploring diffusion-based language models (dLLMs) — LLaDA and Dream framework — evaluated on Hindi/Bengali. Working toward an arXiv preprint on dLLM performance on Indian languages (first-mover territory as of Feb 2026).
Building GigWork-India-Bias — a HuggingFace dataset for gig economy fairness research.
Languages Python · TypeScript · JavaScript · SQL · Rust
AI/ML AWS SageMaker · Bedrock (Claude 3) · LSTM · Isolation Forest · VADER
Backend Node.js · Next.js 14 · FastAPI · REST · Lambda
Data DynamoDB · QLDB · GDELT · Angel One SmartAPI
Infra AWS (Bedrock, SageMaker, QLDB, SNS, Amplify) · Vercel · Stellar
- 🏆 Smart India Hackathon 2025 Finalist — Indigenous Track Monitoring System for Indian Railways
- ⚡ ET Gen AI Hackathon — Shortlisted from ~55,000 teams with ET Sentinel
- 🎯 Targeting Backend AI/ML Engineering roles · Interested in Sarvam AI, AI4Bharat, and Indian-language NLP
- 📖 Deep-diving the Transformer architecture (Attention Is All You Need → BERT → GPT lineage)
- 🔬 Evaluating dLLMs on Indian language benchmarks
- 📚 Building DSA depth in Java alongside ML math foundations