Umama N.

You have a product idea. Or an existing app that needs to see, think, and act. I build the AI layer that makes that happen; fast, production-grade, no agency overhead. I'm a full-stack AI engineer and computer vision specialist. I ship complete products: web, mobile, and AI, using the best tools available in 2026: Claude Code, Cursor, Python, React Native, Node.js, Next.js, NestJS, Ruby on Rails, and purpose-built CV pipelines. Language and framework don't constrain me. The problem does. 𝐖𝐡𝐚𝐭 𝐈 𝐚𝐜𝐭𝐮𝐚𝐥𝐥𝐲 𝐛𝐮𝐢𝐥𝐝: 🧠 Computer Vision Systems I build pipelines that read real human data from standard cameras: no special hardware, no wearables. Pose estimation, gait analysis, movement tracking, emotion recognition, facial verification, object detection, action classification, and real-time anomaly detection. Deployed as web apps, mobile apps, or APIs. 𝐕𝐞𝐫𝐭𝐢𝐜𝐚𝐥𝐬 𝐈'𝐯𝐞 𝐛𝐮𝐢𝐥𝐭 𝐟𝐨𝐫: → Healthcare & Rehab: posture analysis, fall detection, range-of-motion tracking, remote patient monitoring, wound measurement → Sports & Fitness: biomechanical feedback, rep counting, player tracking, injury risk scoring, ball trajectory analysis → EdTech: student attention detection, engagement scoring, gesture interaction, yoga/dance AI coaching, sign language recognition → Finance & KYC: liveness detection, document verification, deepfake detection, behavioral biometrics → Retail: customer heatmaps, shelf monitoring, theft detection, queue analysis, emotion scoring at purchase → Driver Safety: drowsiness detection, distraction monitoring, fleet safety dashboards → Emotion AI: stress detection, pain recognition, burnout monitoring, therapist-support tools → Agriculture: crop disease detection, livestock health monitoring, yield estimation → Construction & Safety: PPE compliance, ergonomic risk scoring, zone intrusion alerts → Security: crowd analysis, intrusion detection, behavioral anomaly flagging 🌐 WebAR (Browser-Based Augmented Reality) AR experiences that run in any smartphone browser: no app download, no headset. Built with WebXR, Three.js, MindAR.js, ARKit, and ARCore. Virtual try-on, body-tracked overlays, marker-based AR for packaging/print, educational AR, and real estate visualization. 📱 Mobile Apps (iOS & Android) Cross-platform and native. React Native, NestJS backends, real-time features (WebSockets, geofencing, live GPS, live bidding). Recently shipped a 40+ screen ride-hailing super app with 8 service modules solo in under 2 weeks, auth to checkout, fully tested and deployed. 🔧 Full-Stack Web & SaaS Platforms End-to-end: React, Next.js, Node.js, NestJS, Ruby on Rails, MERN stack, PostgreSQL, Redis, Docker, AWS. Stripe billing, role-based auth, admin dashboards, real-time collaboration, analytics. Built to scale from day one. 🤖 AI Integration & Automation LLM integration (Claude API, OpenAI GPT-4o, Gemini), RAG pipelines, document intelligence, OCR, semantic search, AI agents, n8n workflow automation. I embed AI where it creates real leverage — not because it's trending. ⚙️ 𝐇𝐨𝐰 𝐈 𝐰𝐨𝐫𝐤: Scope locked and documented before code is written. I use Claude Code and Cursor as force multipliers, what took a team of 4 a month, I ship in days. Milestone-based delivery with async updates every 48 hours. I surface blockers before they become problems, document everything I build, and stay reachable after launch. 𝐈𝐝𝐞𝐚𝐥 𝐟𝐨𝐫: Founders and CTOs who need one accountable engineer to own the full product, from architecture to deployment. If your agency is pushing "phase 2" for features you need now, send me a message. 𝐍𝐨𝐭 𝐢𝐝𝐞𝐚𝐥 𝐟𝐨𝐫: Teams looking to outsource isolated tasks to multiple contractors, or clients who want the cheapest option available. I respond within 2 - 4 hours. Tell me what you're building and I'll tell you exactly what's possible.