Dragon TestingAgentic Testing for Every UI You Ship
Dragon Testing turns requirements into executable UI tests, runs them on real devices, and hands back evidence your team can review — across web, mobile and desktop.
How it works
From Requirements to Reports in Four Steps
Requirement Analysis
Upload requirement documents; the agent parses them into clear user stories, ready for review.
Test Case Generation
The agent writes test cases from those stories, aiming for full coverage of the paths that matter.
Script Generation
Cases become executable scripts with a visual workflow. Pick devices and run them in parallel.
Execution & Reporting
Every run produces a visual report with step screenshots and video replay, so failures are easy to trace.
Product
What the Dragon Testing Agent Does
Describe a test in plain English. The agent plans the interaction path, executes it on a real device, asks when it is unsure, and saves the result as a script you can replay.
Testing Agent
Natural-language test goals become clicks, inputs, swipes and assertions. Start from scratch or extend existing cases step by step.
Human in the Loop
Pause at any step, take over manually or add context. The agent picks up from there and finishes the remaining steps.
Deterministic Replay
Generated tests are saved as scripts. Regression runs replay them directly — stable results, no model calls, near-zero token cost.
Visual Workflows
Test logic is shown as flowcharts you can edit and preview, so complex suites stay readable for the whole team.
Extensible Steps
Add API calls, database queries or custom scripts to any step to cover setup, data checks and back-end verification.
Evidence-Rich Reports
Pass rates, failure reasons, step screenshots and full video replay for every run. Export for audits and sign-off.
See it run
Web Testing Demo
A natural-language flight-search task on Priceline: the testing interface, the generated automation steps, and the saved workflow.
Why Dragon Testing
Built for Teams That Ship Fast
AI generates, scripts execute
The model is used to plan and write tests, not to re-think every run. Once a test passes review it replays deterministically, so results are repeatable and cost stays flat as your suite grows.
Works on the apps you actually have
Web, native mobile, desktop and hybrid apps in one platform, with element-based and visual (OCR / image) locators for screens that resist traditional automation.
Fits AI-generated software
When code is written faster than it can be reviewed, acceptance testing becomes the bottleneck. Dragon Testing checks the result against the requirement, with evidence attached.
Your models, your infrastructure
Run on our cloud or deploy on-premises with the LLM of your choice. Test data and recordings never have to leave your network.
Plans
Pick the Deployment That Fits
Cloud
For individuals and small teams getting started
- AI test generation
- Cloud-hosted devices
- Visual reports and replay
Team
For growing QA and engineering teams
- Full agentic testing
- Unlimited projects
- Register your own devices
- Priority support
Enterprise
For large organizations and regulated industries
- On-premises deployment
- Bring your own LLM
- Custom integrations and SLA
- Source-code licensing available
About
Making Testing Smarter
Dragon Testing builds AI-driven UI test automation. Our first-generation platform combined AI, robotics and model-based testing to automate web, mobile, desktop and hardware-in-the-loop scenarios, and is used by more than a hundred enterprises in finance, rail transit, telecom and enterprise software.
Our second generation puts a large-model testing agent at the center: it reads requirements, writes and runs tests, and turns what it learns into scripts that replay reliably.
We work with customers in the United States and China, and offer both cloud and on-premises deployment.
Dr. Jiangfan Shi
- Ph.D. in Software Testing, University of Nebraska–Lincoln
- 20+ years in testing at IBM, Microsoft and Expedia
- Research cited 1,000+ times; ASE program committee member
- Technical committee, QECon and MTSC testing conferences
Contact
Ready to Get Started?
Tell us what you are testing and we will set up a demo on your own app.
What to include
So we can reply with something useful, tell us: