01 / PERSPECTIVE
Find the overlooked.
Boundaries. Permissions. Exceptions. Use testing viewpoints to explore the risks beyond a successful first attempt.
AI-assisted test design — viewpoints and a matrix you review
See beyond
the happy path.
Turn a spec into a function list, shared viewpoints and a coverage matrix, then into cases you can trace. AI drafts. You review. Kanten does not run the tests.
Kanten sells reviewable test design — functions, viewpoints and cases you can trace and export. It does not run your tests or replace TestRail.
Built for QA teams who review the questions before anyone automates the checks.
Explore the workflowTest AI behaviorSee pricing
100 one-time AI credits for eligible new accounts.
For the people who ask: what could go wrong?PDF SCREENSHOTS TEXT
PRODUCT WALKTHROUGHSAMPLE / ROOM BOOKING
Pick a step. Follow the thinking.
FEATURE / ROOM BOOKING
Illustrative sampleStart with what you know.
Bring a PDF, screenshots or text. Review the extracted requirements before moving on.
room-booking.pdfExample source document
REQ-01
Book an available meeting room
A signed-in employee can book a room with a title, date, start time, end time and attendee count.
- End time must be after start time
- Attendance cannot exceed 6 people
Open question: what if two people book at once?
Interactive sample. No AI calls or project changes.01 / 04
AI proposes. You decide. Test designs are not proof of completeness or execution.
AI BEHAVIOR TESTING · BETA
Define what a good AI answer looks like.
Before changing a prompt, make the expected behavior explicit. Turn business rules into test drafts, then save the criteria you review as a golden dataset.
See the refund-support exampleEXAMPLE: AI CUSTOMER SUPPORT
Start with your business rules.
Refunds are available within 7 days of purchase. After 7 days, refunds are not available. Escalate situations not covered by the policy to a human. Do not invent rules or disclose other customers’ information.
Explore normal requests, boundaries, missing rules and privacy.
ILLUSTRATIVE TEST DRAFTS · NOT MODEL RUN RESULTS
Turn rules into reviewable cases.
| Situation to test | Expected behavior |
|---|---|
| A refund request 3 days after purchase | Explain that the purchase is eligible for a refund. |
| A refund request 8 days after purchase | Explain that the refund window has passed. |
| Exactly day 7, with no cutoff time defined | Flag the unclear boundary. Do not invent a cutoff rule. |
| An exception the policy does not cover | Escalate to a human. Do not promise an unsupported refund. |
| A request for another customer's order | Do not disclose another customer’s information. |
Review once. Keep a baseline you can reuse.
Review the inputs, required behavior and forbidden behavior, then save a version of your test set. Reuse the same criteria when checking answers after a model, prompt or knowledge-base change.
- Align on expected answers
- Explore boundaries and exceptions
- Save and export reviewed criteria
Start designing AI tests
An account is required to save. AI generation uses credits.
Available now: test design, review, saving and export. Generated drafts need human review. Automatic execution against your AI, automated scoring and run-to-run comparisons are not yet available.
Not a test runner
Better tests start with better questions.
Not test automation, and not a test-management clone. For QA engineers and teams who review a design together and leave the thinking visible.
02 / TRACEABILITY
Keep the reasoning.
Review AI drafts, resolve open questions and trace each test case back to its function and viewpoint.
03 / CONTINUITY
Build on your work.
Revisit saved cases and generation batches in your Case repo. Edit, select and export your work to Excel.
Official viewpoint packs
Start from the questions your industry already asks.
Fintech, FIX/FX, SaaS, mobile, online games and performance. Browse the official seeds, then import a pack into your shared viewpoint library and edit it on a project.
Browse viewpoint packs08VIEWPOINTS
Fintech
Payments, balances and regulated account actions. Questions to settle before a money movement is called done.
Browse viewpoint packs08VIEWPOINTS
FIX / FX trading
Order entry on a session such as FIX. Asks whether a trading message can be lost, duplicated or applied twice.
Browse viewpoint packs08VIEWPOINTS
General SaaS
Multi-tenant product work: permissions, drafts, limits and two people editing the same record.
Browse viewpoint packs08VIEWPOINTS
Mobile app
Phone and tablet flows: offline, permissions, the system back control, and a screen that resumes late.
Browse viewpoint packs08VIEWPOINTS
Online game
Live sessions: disconnects, server authority, duplicate rewards, and a spectator who must not play.
Browse viewpoint packs08VIEWPOINTS
Performance
Design questions for slow, large and crowded use. These are reviewable risks, not a load-test runner.
Browse viewpoint packsCurated by Kanten. Import is free and does not unlock a purchase. There is no marketplace and no public comments.
ROOM TO START. SPACE TO GROW.
Start free. Keep your momentum.
Credits are for AI work. Reviewing, editing and exporting to Excel do not use credits.
See all plansFree
GET STARTED$0USD · one-time trial
- 100 one-time AI credits, after you confirm your email address — a trial of the AI features, not a monthly allowance
- One workspace for your own account, with a shared viewpoint library across its projects
- Feature-scoped requirements and reviewed function lists
- Coverage matrix and traceable test-case drafts
Solo
YOUR WORKSPACE$29USD · per month
- 900 AI credits / month
- One workspace for your own account, with a shared viewpoint library across its projects
- Feature-scoped requirements and reviewed function lists
- Coverage matrix and traceable test-case drafts
Team
SHARED WORKSPACE$79USD · per month
- 4,000 AI credits / month, shared by the whole workspace
- Invite colleagues as editors or viewers into one shared workspace — no per-person charge
- Feature-scoped requirements and reviewed function lists
- Coverage matrix and traceable test-case drafts
Do I need test-design experience?
No. Each step explains what it is for and what a good row looks like, and the shared library already holds the perspectives experienced testers reach for — boundaries, permissions, interrupted flows — so you start from questions rather than a blank page. If you are experienced, you can override every AI decision, write your own viewpoints, and keep several concrete variations per function and viewpoint.
Can AI miss issues?
Yes. AI proposes; it does not certify, and it can be wrong, incomplete, or label a row wrongly. Kanten is built around that: prompts ask the model to mark source statements, assumptions and open questions, the function list and the viewpoint review cannot be skipped, and what the workbook records is your decisions and the questions still unresolved. Treat every draft as something to review, not as evidence of coverage.
Is this test automation?
No. This version does not run the browser and does not generate scripts. It is analysis, a repo of functions, viewpoints and cases, and Excel export.
What spends AI credits?
Only model work: drafting a function list from the material you provide, the AI coverage review, marking the function × viewpoint matrix, and writing cases. Kanten shows a planned estimate before a run and charges the usage the run actually reports — the estimate is a plan, not a guaranteed maximum. Reviewing, editing by hand, applying a draft and exporting cost nothing.
Can I reuse viewpoints?
That is the point of the shared library. One library serves every project in the workspace, each viewpoint carrying where it applies, so the next feature inherits the question instead of rediscovering it. You can edit or add viewpoints at any time, free of charge.
YOUR NEXT FEATURE. A CLEARER PERSPECTIVE.
Make the test decision visible.
Try demo — no accountNo account for the sample. Create one when you want to keep a project or export Excel.