The future we’re building
Software changes quickly. Reliable testing needs context, judgment, and evidence. We’re developing bugAgent into a teammate that can discuss the product, do the testing work, and explain what it found.
The core quality platform is available today. Conversational testing is in active development; not yet released. The complete autonomous workflow remains in research and development.
The platform and the path ahead
Start with the quality tools available now. Follow the work toward a tester that can understand, investigate, and verify alongside your team.
Available today
Connect bug reports, test cases, automation, AI-assisted analysis, and human testing expertise.
Features vary by plan. Selected capabilities have their own beta or pilot status.
Explore current featuresIn active development; not yet released
TestLauncher Live is the conversational testing experience we’re developing inside bugAgent.
Private prototypes are being evaluated. This experience is not generally available.
Explore conversational testingResearch and development
Our longer-term goal is a tester that carries approved intent through testing, investigation, and verification.
The complete workflow is not yet released. Research and validation will determine its scope.
Explore the research directionConversational testing · In active development; not yet released
Invite a project-bound agent to a walkthrough. Explain a change, demonstrate a problem, or ask a question. Our private prototypes are exploring two-way voice, screen sharing, and scoped project questions inside bugAgent.
The goal is natural collaboration: answer when addressed, listen when others are talking, and stop speaking when interrupted. Background testing work should let the conversation continue.
Illustrative future conversation
“This workflow must still work when the connection drops.”
“Should changes sync automatically when the connection returns, or should the user confirm?”
“Sync automatically, without creating duplicates.”
“I’ll propose tests for reconnection and duplicate prevention for your review.”
A product concept, not a transcript or a claim that the complete workflow is available.
Autonomous tester · Research and development
We’re researching the complete workflow below. Its value depends on useful defect detection, reliable verification, and fewer unnecessary handoffs—not simply generating more scripts.
Connect approved requirements, project knowledge, and the build under test. Identify gaps and ask questions before deciding what correct behaviour means.
Propose test cases and Playwright scripts tied to the team’s intent. Keep the testing plan reviewable before execution.
When a system is hard to test, propose fixtures, diagnostic tools, or a test harness. Validate the helper before trusting its results.
Distinguish an application defect from a broken test, an environment problem, or an inconclusive result. Show the supporting evidence.
Propose locator and script repairs that preserve the expected behaviour. Check them independently so a passing test does not conceal a defect.
Explain what was tested, what failed, and what remains uncertain. People retain authority over application changes and release decisions.
We will evaluate these methods against existing tools, retain failures, and seek independent validation. The autonomous tester thesis has not yet been proved, and no release date is promised.
How teams could work
The intended experience connects product, engineering, and QA. A team can start with a discussion or assign a bounded testing task directly in the project.
Requirements guiding the research
Control over data, actions, and participation is part of the product direction.
Keep each session and delegated worker within its workspace and project permissions. The current guest pilot does not provide access to project records.
Require participant consent, host-controlled guest admission, and visible controls to stop agent work, microphone use, and screen sharing.
Evaluate local masking before permitted cloud processing, with recording off by default in the pilot. Detection can miss sensitive information; masking is not a guarantee.
Work within an approved scope and budget. Escalate uncertain results, proposed changes, and release decisions to the people responsible.
Yes. The core platform supports bug reporting, test management, automation, and integrations. Features vary by plan, and selected capabilities are labelled as beta or pilot. Explore current features or book a discovery call.
No. TestLauncher Live is in active development; not yet released. Private voice and screen-sharing prototypes help us evaluate the experience. They are not a generally available meeting assistant.
Native desktop and mobile clients, Canadian French, phone participation, and customer-hosted or offline configurations are future directions to evaluate. A dedicated StealthOS testing environment is a research possibility, not an available bugAgent product.
The current production database is in US East. Canadian data residency is a planned option for new clients, not active in production today. Storage, AI processing, media, backups, and subprocessors must be scoped for each engagement. Read the Trust Center.
Built by TestLauncher
We’re a Canadian-owned software innovation company building for teams where reliability, control, and integration matter. Tell us about your application and the testing work that is hardest to do well.
Book a discovery call