█████╗ ██╗ ██╗ █████╗ ███████╗██╗ ██╗ ███╗ ███╗███████╗ ██╔══██╗██║ ██╔╝██╔══██╗██╔════╝██║ ██║ ████╗ ████║██╔════╝ ███████║█████╔╝ ███████║███████╗███████║ ██╔████╔██║███████╗ ██╔══██║██╔═██╗ ██╔══██║╚════██║██╔══██║ ██║╚██╔╝██║╚════██║ ██║ ██║██║ ██╗██║ ██║███████║██║ ██║ ██║ ╚═╝ ██║███████║
Paste a Job Description — I'll match it against my skills, highlight relevant projects, and generate an ATS-optimized summary.
I build explainable AI systems that solve real business problems — from fraud detection to credit risk analytics. 10+ live deployed applications, certified by Oracle, Google, Meta & AWS.
I'm a Computer Science undergraduate specialising in Data Science, focused on building AI systems that are not just accurate — but explainable. My work spans fraud detection, credit risk analysis, multilingual speech systems, and real-time collaborative platforms.
I specialize in Explainable AI (XAI) using SHAP — translating ML predictions into plain-English business decisions that compliance teams, HR managers, and analysts can actually act on. This isn't a side skill; it's the core thesis of every project I build.
Currently interning at PetOlife Technologies as a Backend Developer, working directly with the CTO. Actively targeting AI startups, fintech, and product companies for full-time roles. Also learning German 🇩🇪 on the side.
Production-grade multilingual Speech-to-Text platform. React/TypeScript frontend + FastAPI backend deployed on Vercel. Groq Whisper large-v3 for ASR with LLaMA 3.3-70b summarization layer.
Production-grade Kanban board where multiple users collaborate simultaneously. Node.js + Socket.IO + PostgreSQL backend with real-time conflict resolution. AI task suggestions via Groq API.
SHAP-powered credit risk decisions + bias detection + LLM counterfactual appeal system. "Your loan was denied because X — here's what to change." Built with FastAPI + React, deployed on Vercel.
Cost-sensitive ML pipeline where FN penalty = 5× FP — mirroring real banking loss ratios. XGBoost + SHAP TreeExplainer + FastAPI. Real-time transaction scoring with full explanation per prediction.
NLP + ML pipeline classifying job postings as legitimate or fraudulent. Paste any JD — instant risk score with SHAP attribution showing exactly which phrases triggered the fraud flag.
Random Forest + SHAP — tells HR not just WHO will leave but exactly WHY. Per-employee SHAP waterfall charts.
🥇 FUSIONX 2026 — Best AI Innovation for Social Impact. Zero-backend AI farming app with 7 features: crop advisory, yield forecasting, disease detection, live market prices. Voice-enabled in 6 Indian languages. Built in 48 hours, works on ₹2000 phones.
NLP + ML system to detect fraudulent medical certificates — addressing a real societal problem of medical fraud in employment and insurance contexts. 3rd place at Ideathon 2026.
TF-IDF + semantic similarity engine accepting ANY dataset. Handles cold-start users and produces per-item natural-language explanations. Works for movies, jobs, products — no retraining required.
AI-powered BI system combining XGBoost + Random Forest + LLM reasoning to forecast trends, detect anomalies, and generate actionable business insights from any dataset.
Rigorous A/B testing using hypothesis testing, confidence intervals, and p-value evaluation to drive data-backed decisions on conversion uplift. Production-realistic statistical analysis.
Flask web app using SpeechRecognition API supporting multiple Indian languages with real-time audio processing. Direct production deliverable from Hostingsignal internship.