Home/Workbenches/Vision / Computer Use

Vision / Computer Use

AI-powered screen automation, OCR, GUI interaction, and visual testing — replacing brittle RPA with intelligent computer use agents.

Launch Workbench View Capabilities

Visual Intelligence

AI that sees and acts.

Vision models that understand screens, documents, and interfaces — then take action.

🖥️

Screen Understanding

AI models that comprehend UI layouts, text, icons, and interactive elements in real time.

🔤

Advanced OCR

Multi-language text extraction from documents, screenshots, and video frames with layout preservation.

🤖

GUI Automation

Autonomous agents that navigate desktop and web applications using visual understanding.

🧪

Visual Testing

Automated visual regression, accessibility audits, and cross-browser comparison testing.

📋

Process Recording

Record human workflows, extract steps, and generate reproducible automation scripts.

🔄

RPA Migration

Convert brittle selector-based RPA bots to resilient vision-based automation agents.

95%+

OCR accuracy

10x

More resilient than RPA

50+

App integrations

24/7

Unattended execution

Workflow

Record. Understand. Automate.

Record

Capture human interaction with applications — clicks, typing, navigation patterns.

Understand

AI analyzes recordings to identify intent, decision points, and error handling.

Generate

Produce vision-based automation agents that adapt to UI changes.

Scale

Deploy agents across environments with monitoring, error recovery, and audit trails.

Universal Workbench Capabilities

Available in the Nebula Cloud Universal AI Workbench

These capabilities ship inside the two Universal Workbench editions. Choose an edition and launch — no separate product to buy.

Computer Vision

Detection, segmentation, and 3D perception.

Application	Availability	Licence
Kornia	Ubuntu	Built-in
ONNX Runtime GPU	Ubuntu	Built-in
Open3D	Ubuntu	Built-in
YOLO	Ubuntu latest	Built-in

Configure the Ubuntu 24.04 Universal Workbench