Test with AI
Manual and exploratory testing services, with AI doing the preparation
Some defects only a person finds: the confusing flow, the wrong wording, the screen that works but feels broken. You get a tester on your next release, bug reports with steps and evidence in your own tracker, and a written view of what to keep testing by hand and what to automate.
- Change the booking date after payment Price, confirmation email and calendar entry stay consistent bug filed
- Use the form on a narrow phone screen Every field reachable and readable ok
- Error message after a declined card Wording is clear to a first-time user tester
- Tester-invented path, back button mid-checkout No duplicate order is created ok
- Since 2015
- 4.9 out of 5 on Clutch
- 50+ testimonials
At a glance
- What it is
- Manual and exploratory software testing
- First step
- Trial test About one week
- You receive
- Bug reports in your tracker with steps and evidence +2 more
- After that
- Pilot Two to six weeks
What we test
Services in this area
-
Exploratory testing
Time-boxed sessions with a written charter, where a tester learns the product while testing it and follows what looks wrong.
- A charter per session: what to explore and what risk to look for
- Session notes that show what was covered and what was not
- Bugs with steps, evidence and a severity you can act on
-
Release checks by hand
Planned checks of the paths that must work before each release, run by a tester who knows your product.
- Checklists kept current as the product changes
- Checks that repeat every release flagged for automation
- Release notes that state the open risks
-
Test design and test cases
Test cases written from requirements and from the product itself. AI proposes ideas; the tester cuts them down.
- Edge and negative cases proposed by AI, selected by a tester
- Cases tied to requirements so gaps are visible
- Only cases that will be run are kept
-
Usability, device and compatibility checks
Checks on the browsers, devices and screen sizes your users have, and on whether the product makes sense to them.
- Browser and device matrix agreed from your own usage data
- Findings on layout, text and flow, not only on crashes
- Accessibility problems a person notices on the page
Why manual testing is still needed
Automated tests check what someone already thought of. A tester finds what nobody thought of.
- New features have no scripts yet. Before a feature is stable, writing automation costs more than testing it by hand.
- Some defects are not assertions. A confusing label, a flow that technically works or a layout that breaks on one phone is hard to script and easy for a person to spot.
- Exploration finds the unexpected. A tester who follows an odd result often finds a second defect behind the first.
The slow part of manual testing is usually not the testing. It is the preparation: working out what to test, getting data into a usable state and writing up what was found. That is where AI helps, so the tester’s time goes to using the product.
How a session works
- Charter. We agree what to explore and which risks matter, in a few lines.
- Preparation. AI proposes test ideas, edge cases and test data from your requirements. The tester removes what is wrong or pointless and keeps the rest.
- Testing. The tester uses the product, takes notes and follows what looks wrong.
- Reporting. AI drafts bug reports and a session summary from the notes. The tester corrects them and files them.
- Review. We go through the findings with you, including what was not covered.
What a session looks at
| Area | Example checks |
|---|---|
| Core flows | The paths your users take every day work from start to end |
| Edges | Empty, very long, unusual or wrong input; interrupted steps; back button; double click |
| Consistency | The same fact shown in the page, the email and the invoice agrees |
| Wording and layout | Messages are clear, text fits, nothing overlaps on a small screen |
| Change impact | What a new change could have broken nearby |
When checks start to repeat on every release, they belong in automated regression testing; scripted functional checks are covered by automated functional testing, and the full range by QA automation testing services.
How an engagement runs
Start small, with a fixed scope
Each step has a defined scope, duration and deliverables. You decide after each one whether to continue.
- Step 1 About one week
Trial test
One release or feature is tested by a Zapple.Tech engineer in charter-based sessions. You see how defects are found, described and reported before you commit to more.
You get
- Bug reports in your tracker with steps and evidence
- Session notes and a summary of what was and was not covered
- Written view on what to test by hand and what to automate
- Step 2 Two to six weeks
Pilot
A tester, or a small group, works in one product area through several of your releases, with checklists, test data and reporting set up.
You get
- Test cases and checklists for the area
- Test data sets drafted with AI and checked by the tester
- A short report after each release
- Step 3 Monthly
Ongoing
Testers become a regular part of your release cycle, and checks that repeat move to automated scripts written by the same team.
You get
- Testing for each release on a schedule you agree
- A list of checks moved to automation, kept up to date
- Monthly summary of defects found by area and severity
FAQ
Questions buyers ask
What is the difference between manual and exploratory testing?
Manual testing means a person runs the tests. Scripted manual testing follows prepared steps. Exploratory testing has the tester design and run tests at the same time, guided by a charter. Most releases need some of each, and the mix is set by product risk.
Where exactly is AI used?
Before and after the testing. Before: proposing test ideas, edge cases and test data, and turning requirements into checklists. After: drafting bug reports and session summaries from the tester's notes. During the testing the tester decides, and the tester reads and corrects every AI draft before you see it.
Who does the testing?
Zapple.Tech QA engineers, who also write test automation. That is why repeated checks are moved to scripts instead of being run by hand forever.
Why not automate everything?
Automation checks what someone already thought of, and it costs more than it saves on a feature that is still changing. Exploratory testing finds what nobody thought of. A good split uses both, and the trial report says where your split should be.
What do you need from us?
Access to a test or staging environment, test accounts, access to your bug tracker, and a short description of the release. Requirements or tickets for the features under test help the tester prepare.
Which devices and browsers do you cover?
The device and browser matrix is agreed from your own usage data and written into the quote, including any devices it does not cover.
How long does it take?
The trial covers one release in about a week. The pilot runs two to six weeks, long enough to go through at least two of your releases.
Will our data be sent to an AI service?
Not without your agreement. What may be used with AI tools is agreed before the start, and requirements, test data or notes that contain personal data stay out of AI tools unless you say otherwise.
How much does it cost?
Prices are not published. The quote depends on how often you release, how many product areas are in scope and the size of the device matrix. Tell us those and you get a scoped quote within one business day.
Start with a one-week exploratory test of one release
Tell us which release or feature needs a second pair of eyes, and how often you ship. We reply within one business day with scope, timeline and a quote.
- On the market since 2015
- Rated 4.9 out of 5 on Clutch, from 7 client reviews
- More than 50 testimonials
Prefer to talk first? Schedule a call