My Resume

Professional Background & Experience

An overview of my experience, projects, and technical proficiencies — focused on transforming perception into action through real-time vision systems, embedded control, and edge AI optimization.

Umer Ghafoor

Computer Vision & Robotics Engineer

Focused on transforming perception into action through real-time vision systems, embedded control, and edge AI optimization, building scalable robotics solutions that operate in dynamic real-world environments.

umerghaforr@gmail.com
+923016339325
linkedin.com/in/umerghafoor
github.com/umerghafoor
umerghafoor.github.io

Experience

Robotics & Computer Vision Engineer

BeetleOps

November 2025 – Present

  • Leading an AI team and acting as system architect for an end-to-end computer vision product built from scratch.
  • Developed real-time CV pipelines for edge devices (detection, tracking, low-latency inference) using GStreamer (NVDEC) and optimized deployment with TensorRT and other model optimization techniques on embedded systems.
  • Optimized end-to-end pipeline latency through async processing and multi-threaded inference execution.
  • Achieved real-time inference under strict latency constraints via GPU utilization and pipeline optimization.
  • Developed a RAG-based log analysis system for intelligent system log debugging and insights.
  • Designed WebRTC streaming with STUN/TURN servers, including VPS-hosted custom STUN/TURN infrastructure for NAT traversal.
  • Built dashboards using Next.js and React Native for real-time monitoring and visualization.
  • Managed remote deployments for USA-based edge systems using Tailscale, RTSP & Nmap.

AI Intern

CHI Technologies

June 2025 – August 2025

  • Worked on Angular and Node.js based web applications for healthcare automation systems.
  • Integrated speech-to-text pipelines using AWS, Whisper, and Speechmatics for real-time transcription workflows.
  • Built automated SOAP note generation and real-time doctor–patient interaction systems.
  • Developed a real-time doctor–AI interaction system using WebSockets (Socket.io) for low-latency communication and AI agent event streaming.

Projects

Precision Farming — Autonomous Weed Detection & Targeting Robot

  • Built a robotic system using Raspberry Pi for real-time weed detection and targeting in dynamic environments.
  • Implemented YOLO-based object detection for identifying weeds from live camera streams.
  • Designed a ROS2-based modular architecture, separating perception, decision-making, and actuation into independent system nodes.
  • Implemented actuation logic using a laser-based targeting mechanism, controlled through a desktop interface with real-time system communication.
  • Developed a Qt (C++) desktop application with widget-based UI, featuring a digital robot simulation integrated into the interface for system visualization and control.
  • Enabled network-based communication over Tailscale, allowing secure remote control and data exchange between the robot and the desktop application.

IoT Plant Monitoring System

  • Developed an ESP32-based system to monitor humidity, temperature, and soil moisture.
  • Sends real-time data to AWS IoT Core for cloud-based monitoring.

Technical Proficiencies

Robotics, Computer Vision & AI

Robotic Systems

Raspberry Pi, STM32, ESP32, Arduino, NVIDIA Jetson (Orin Nano), ROS2.

IoT & Cloud

MQTT, RTSP, AWS IoT Core, Real-Time Monitoring, Sensor Integration.

Computer Vision

TensorRT (Jetson optimization), ONNX, FFmpeg & GStreamer (video streaming pipelines), model optimization (quantization, pruning, distillation), real-time computer vision systems (detection, tracking, low-latency inference), LLM fine-tuning.

Supporting Technical Skills

Software Development

PyQt6, C++, Java, Next.js, React Native, Flutter, Qt Framework.

Networking & Systems

WebRTC (Real-Time Communication), TCP/IP Networking, RTSP, Tailscale, Git, GitHub, GitLab, Linux, Agile Development.

3D Design & Tools

SolidWorks, Fusion 360, AutoCAD, Blender, Adobe Creative Suite.

Education

BS in Computer Science (Robotics & Automation)

FAST – National University of Computer and Emerging Sciences, Islamabad

Relevant coursework: Data Structures & Algorithms, Object-Oriented Programming, Operating Systems, Computer Networks, Database Systems, Artificial Intelligence, Machine Learning for Robotics, Digital Image Processing, Robotics Technology, Embedded Control Systems, Software Engineering, Parallel and Distributed Computing, Information Security, Generative AI, Statistical Modelling, Discrete Structures, Linear Algebra, Probability and Statistics.

Awards & Achievements

Dean's List of Honor

FAST-NUCES

2022

Certifications

Machine Learning Specialization

Issued: Oct 2023

OpenCV Bootcamp

Issued: Jul 2023

TensorFlow Keras Bootcamp

Issued: Jul 2023

Extracurricular

Co-Founder & General Secretary

FAST-LADS (Leaders Advancement & Development Society)

Coordinator, Graphic Design

IEEE-FAST & ISYWSC 2022

2022 – 2023