I turn AI models into real, deployed products — not notebooks. Currently a CSE undergrad shipping full-stack AI platforms, computer vision systems, and NLP engines end-to-end.
🛠️ Tech Stack
Languages
Frontend
Backend
AI / ML
Databases & Cloud
🚀 Featured Projects 🗂️ JobTracker AI — Full-Stack AI Job Application Platform
Problem: Job seekers lose track of applications and can't tell if their resume actually matches a role. Solution: Full-stack platform with an interactive Kanban board and AI-powered resume analysis, deployed to production with a mobile companion app.
⚙️ Built with Next.js, TypeScript, PostgreSQL, NextAuth.js (Google OAuth) — 6-stage Kanban pipeline + 6 REST API endpoints 🤖 Integrated GPT-4o-mini for two AI features: a Job Description Parser (extracts 12+ structured fields) and a Resume Match Scorer (score, A–F grade, missing skills, ATS keywords) 📱 Shipped to Vercel with an Android APK (Capacitor) for mobile access 🔗 Live Demo · Code 🕵️ Deepfake Detection System — Computer Vision & Transfer Learning
Problem: Manually verifying image authenticity doesn't scale. Solution: A fine-tuned deep learning classifier served through a low-latency production API.
🧠 Fine-tuned ResNet-50 on 4,000+ labeled images — 87.25% accuracy, 0.9424 AUC-ROC ⚡ Built a FastAPI inference backend optimized for low latency, containerized and deployed on Hugging Face Spaces 🔗 Live Demo · Code 🎬 CineMatch — NLP Recommendation Engine
Problem: Users need content-based movie recommendations without relying on collaborative filtering data. Solution: An NLP-driven recommender served via a real-time REST API.
📊 Processed 4,800+ movie records, using NLTK tokenization and stop-word removal, cosine similarity for content-based matching 🚀 Deployed on Render with Flask, integrated the OMDB API for sub-second recommendation retrieval 🔗 Live Demo · Code
📊 GitHub Stats & Activity
🏆 Honors & Recognition SCROLLS 2025 — Winner, Technical Research Paper Competition (AI/ML methodologies) National AI Agent Product Space Hackathon — Designed and deployed an LLM orchestrator with deterministic fallback mechanisms GSSoC 2026 — Open Source Contributor
🤝 Let's Connect
I'm currently looking for full-stack AI/ML internship opportunities where I can ship production systems, not just prototypes. If you're building something interesting — or just want to talk shop about RAG pipelines, model deployment, or Next.js architecture — my inbox is open.
📫 shrisha337@gmail.com | 💼 LinkedIn | 🌐 Portfolio — dropping soon