Ubaid ur Rehman
AI Engineer + ML Specialist, Computer Vision, LLM Developer, Document AI, and Automation
GLM-OCR Document Intelligence Pipeline
Production-grade bilingual (Arabic & English) OCR system using a 1.1B-parameter GLM-OCR model. Achieves 69% accuracy with 1.2s inference latency, integrated with Azure Document AI at 10Pearls.
YOLOv11 Real-Time Object Detection
High-performance real-time object detection system built with YOLOv11 and PyTorch. Supports multi-class detection at 40+ FPS with a custom training pipeline, deployed for surveillance and industrial inspection use cases.
LLM-Powered Document Automation Platform
AI automation platform leveraging GPT-4 and custom LLM agents for intelligent document processing — extracting, validating, and structuring invoice & form data into structured JSON outputs via agentic function-calling pipelines.
Hi, I'm Ubaid
I'm Ubaid ur Rehman, an AI Engineer & ML Specialist based in Pakistan. I specialize in Computer Vision, Document Intelligence, and Large Language Models (LLMs). My work includes building production-grade OCR pipelines (GLM-OCR, 1.1B params), YOLO-based object detection systems, and Azure Document AI integrations at 10Pearls and FormMate.
I'm passionate about pushing the boundaries of AI in real-world applications — from 69% GLM-OCR accuracy improvements to 1.2s inference latency at scale. I'm always open to exciting AI/ML projects and collaborations, so feel free to reach out.
Send me a message