Local AI Image Processing Engineer (GPU / Desktop)

FabTechSol; Pakistan; On-site; PKR 1,800,000 - 3,000,000; Full time

Job summary

FabTechSol is seeking an experienced AI Image Processing Engineer to build a GPU-accelerated Windows desktop application for local use on NVIDIA RTX hardware.

You will capture images from a connected camera, composite subjects into themed backgrounds, and render the final output offline in 5–10 seconds with QR codes generated and stored locally, ensuring premium quality output.

Qualifications

• Experience building GPU-accelerated desktop apps on Windows.
• Strong background in local image processing pipelines for real-time performance.
• Proficiency with ControlNet, CUDA and neural inference on RTX hardware.

Responsibilities

• Develop a full-screen Windows desktop app with camera capture and local rendering.
• Integrate ControlNet into local inference pipelines and optimize performance.
• Implement offline rendering with 5–10 second turnaround and local QR code generation.

Skills

ControlNet integration

CUDA / GPU acceleration

TensorRT / ONNX Runtime

VRAM management

Electron

React-based UI

Camera integration

QR code generation

Local file management

Image compositing

Lighting matching

Watermark overlays

Tools

Electron

React

Job description

Local AI Image Processing Engineer GPU / Desktop

FabTechSol is looking for an experienced AI engineer to join our team for an exciting client project involving a GPU-accelerated AI image processing desktop application running locally on Windows with NVIDIA RTX hardware.

Project Overview

What you'll be building

You'll be building a full-screen Windows desktop application that:

• Captures images from a connected camera
• Composites subjects into themed backgrounds
• Renders final output in 5-10 seconds - offline, no cloud dependency
• Generates QR codes and stores images locally

This is a premium quality output application (social sharing + print-ready), not a novelty filter tool.

Required Skills

AI / ML

• ControlNet integration in local inference pipelines
• TensorRT and/or ONNX Runtime for inference optimization
• VRAM management across continuous long-duration sessions (6+ hours)
• CUDA / GPU acceleration on RTX hardware

Frontend / Desktop

• Electron or React-based desktop UI
• Camera hardware integration
• QR code generation and local file management

Image Processing

• Clean compositing and masking
• Lighting matching for professional-grade output
• Toggleable watermark/branding overlays