We build AI test automation that business teams own.
AutoTestX is an early-stage company based in Canada. We help teams stop re-testing every release and upgrade by hand, with a dedicated edition for IBM Maximo.
From UAT spreadsheets to tests that run themselves.
- The problem
Regression testing that never got easier
Business teams re-tested every release, patch and upgrade by hand, from UAT spreadsheets, with screenshots pasted into Word as evidence. Generic test recorders struggled with enterprise screens (lookups, frames, table rows, generated ids), and scripted tools needed developers most teams did not have.
- Building
Built on one of the hardest web applications to test
We built AutoTestX against real IBM Maximo 7.6 flows first: purchase orders, GL account lookups, inventory and assets. If it works there, it works on most business applications. The recorder finds fields by stable IDs first, table rows are matched by business key, and Maximo’s own messages become checks. Tests are steps an analyst can read, not code.
- Building
Evidence on every run
Every run produces a report with each step, the data used and a screenshot of the screen. Steps that find their field by a fallback are flagged, so teams hear about UI drift before tests fail. A small agent lets testers record behind their own VPN.
- September 2026
The hosted platform goes live
AutoTestX opened as a hosted service at app.autotestx.com, on Amazon Web Services in Canada, with HTTPS throughout and daily backups.
- October 2026
Website, documentation and design partners
autotestx.com and the full product documentation went public, and we opened three places for design partners: a free 60-day pilot on ten of their own test cases.
- October 2026
One platform, and a Maximo edition
AutoTestX opened to teams testing any business web application, with AI drafting, healing and diagnosis across the platform, and the Maximo-specific support kept as a dedicated edition.
- October 2026
Self-healing, schedules, a Mac agent and AI diagnosis
Shipped since launch: self-healing with suggested fixes, rules and an audit trail; scheduled suite runs; CI integrations; the agent for macOS; and optional AI diagnosis of failures, on our AI key or the organization’s own.
- Next
Built with our first customers
The roadmap is set by the teams using it: single sign-on and retention settings are next in line. The capability list on the product page says what is built and what is not, release by release.
Three things we hold ourselves to.
- 01
Analysts own the tests
The people who know the business process write and read the tests. Developers are welcome, not required.
- 02
Evidence over assurances
Every run leaves a record a change board or an auditor can read without us in the room.
- 03
Say what is missing
We publish what the product does not do yet, so nobody discovers it halfway through a pilot.
Talk to the people building AutoTestX.
In 30 minutes: a live demo of AI drafting, healing and explaining tests on a representative flow, and what your team would need for a pilot.