n
nexariza

Nexariza Ai

@nexariza

AI Automation, OpenAI Claude Integration, Custom AI Chatbots, Agentic AI

Pakistan
Engels, Arabisch
Sommige informatie wordt in het Engels weergegeven.
Over mij
We are Nexariza AI—a 7-person AI development agency from Pakistan specializing in agentic AI systems, business automation, and computer vision. Founded by Ahmad Yasin, our team has deployed 60+ production AI systems for clients across the US, EU, and MENA region. Our delivered projects include a multimodal AI surveillance platform, a face recognition attendance system deployed on real industrial hardware, a medical AI diagnostic model achieving 0.92 AUC, and Detectra AI — our award-winning final-year project. We build with Claude Code, LangGraph, VAPI, n8n, YOLOv8, FastAPI, React, and Next.js.... Lees meer

Skills

n
nexariza
Nexariza Ai
offline • 
Gemiddelde reactietijd: 1 uur

Bekijk mijn diensten

AI-technologie advies
I will build a yolov8 object detection or face recognition system in python
AI websites & software
I will build a face recognition attendance system for your school or office

Portfolio

Werkervaring

Nexariza_AI

CEO & Founder

Nexariza AI • Fulltime

Jul 2024 - Present2 yrs 1 mo

* Founded Nexariza, a specialist AI/ML consultancy delivering production-grade systems for US, EU, and MENA clients - computer vision pipelines, agentic LLM products, medical AI diagnostics, and full-stack SaaS. * Architected Detectra AI (company flagship): multimodal video intelligence platform fusing YOLOv8, ByteTrack, Whisper STT, and EasyOCR via CrossModalTransformer into a real-time surveillance pipeline with RAG-ready JSON output. * Delivered a production embedded face recognition system (SCRFD + AdaFace + ONNX Runtime) on industrial Ubuntu 22.04 hardware - RTSP ingestion, SQLite face database, GPIO/Wiegand-48 access control; eliminated 100% of client's manual access overhead. * Designed and deployed 10+ full-stack AI SaaS products (Next.js . FastAPI . Docker) on Neon PostgreSQL, Upstash Redis, and Cloudflare R2; led a 5-person engineering team with CI/CD via GitHub Actions.