Test with AI

Functional testing services: features checked against requirements, automated and manual

If features reach a release without anyone checking them against what was asked for, and your users are the ones who find the gap, our functional testing services and automated functional testing services close it. Every new or changed feature is checked against its requirements before it ships, so a feature that does not do what its story says is found by a tester rather than by your users, and a missing rule becomes a question in the sprint instead of a support ticket after the release. Testers read your stories while they are written and turn each acceptance criterion into a test case; checks that repeat become automated functional tests in your repository and CI. AI drafts test ideas and tests, and a tester or engineer corrects and reviews every draft. We start with a one-week audit that maps your requirements against the tests you have.

Functional audit, requirements mapped to tests 1 passed 1 failed 2 review
  • Sign-in with a valid account Criterion is written, the steps are stable, the check repeats every release automate
  • Discount code at checkout The requirement does not say what happens to an expired code, so a question goes to the product owner question
  • New onboarding screen Still changing every sprint, so a tester checks it by hand first manual first
  • Report export to a file No requirement written yet, so no test is written until one is agreed gap
Illustrative example of an audit worksheet. Not client data.
  • Founders of CodeceptJS and Codeception
  • Founders of Testomat.io and Testomat.ai
  • 32 test automation engineers
  • 42 completed projects
  • Since 2015
  • 4.9 out of 5 on Clutch

At a glance

What it is
Functional testing
First step
Functional test audit About one week
You receive
Requirements and stories mapped against existing tests +4 more
After that
Pilot Two to six weeks

What we test

Services in this area

  • Feature testing in the sprint

    Testers read stories while they are written, turn each acceptance criterion into a test case and check the feature when it is built.

    • Gaps in a requirement raised as questions, not guessed
    • Test cases kept in your tracker or test management tool
  • Automated functional tests in CI

    Checks that repeat on stable features become automated tests in your repository, run in your CI.

    • Smoke checks on each build
    • Tests drafted with AI agents and reviewed by an engineer before merge
  • System testing before a release

    Full business flows run through the whole product before it ships, so a problem between features shows up before your users find it.

    • Flows chosen with you from the audit ranking
    • A written result after every release
  • Manual functional checks by testers

    New and changing features, and anything that needs a person's judgement, are checked by testers first. They are scripted only once they settle.

    • Exploratory and manual testing is described on its own page

Functional testing services: what we check

We check that every new or changed feature does what its requirements say, release after release. That covers the screens a user sees and the rules behind them: a discount code that applies, a form that rejects a bad entry, a role that can see only what it should. The work is mostly automated tests in your repository and CI, with testers where a person’s judgement counts.

Re-checking features that already worked is regression testing. Checking how your services and third-party systems work together is integration testing. This page is about the feature in front of the team this sprint. Whichever functional testing company you work with, ask to see how one of your stories would become a test case; it shows quickly whether the vendor reads requirements or only clicks through screens.

From requirements to test cases

Testers read your stories and acceptance criteria while they are written, not after the feature is built. Each criterion becomes a test case: a step, the data, the expected result.

When a requirement is unclear, we do not guess. We raise the gap as a question to the product owner. The answer goes into the story, and the test case follows from it. Test cases live in your tracker or test management tool, so they stay yours.

What a functional test pass covers

  • Feature testing each sprint: the new or changed feature, checked against its acceptance criteria.
  • Smoke checks on each build: a few fast tests that say whether the build is worth testing further.
  • System testing before a release: full business flows run through the whole product, to find problems that appear only between features.
  • Support for acceptance checks: your business users run the final check; we prepare the checks and the data. See user acceptance testing. The decision to accept stays with you.

Automated functional testing: what we automate and what we do not

Checks of stable features that repeat every release are automated, in your repository and CI. New and changing features are checked by testers first. Once a feature settles, the same team moves the repeated manual checks to scripts.

The reason is cost. A test written against a feature that changes every sprint breaks every sprint, and the team stops trusting it. The audit marks which features are stable enough to automate now and which wait. Checks that are manual by nature are described under manual and exploratory testing.

On web, mobile, API and desktop

Functional testing applies on every platform. For the automation side, see:

Where AI helps, and where people decide

AI drafts test ideas and test cases from your stories. A tester corrects every draft before you see it, and personal data stays out of AI tools by default. For automated tests, agents draft new tests and propose repairs when a screen changes. An engineer reviews each change before it is merged. Agents work only against your test environments, and the environments, the data and the AI provider are agreed first.

An AI draft does not know your business rules. It can restate a requirement slightly wrong and still read well, which is why a person checks each one. It also cannot tell you that a requirement is missing; that comes from a tester reading the story.

How the tests are written

Zapple.Tech was founded by the founders of CodeceptJS and Codeception, open-source testing frameworks. Codeception runs acceptance, functional and unit tests in one framework, and CodeceptJS was designed for scenario-driven acceptance testing. Where your framework allows it, we write functional tests in the same spirit: one requirement, one scenario, in steps a product owner can read. If your team already uses Codeception, see Codeception support.

When functional testing will not help yet

Functional testing needs requirements to check against. If nobody can say what a feature should do, tests cannot settle it; the first step is to write that down, and a tester can help with the questions. If the product changes shape every week, test cases go stale before they are run. And if the complaint is slowness or a security concern, the right service is performance testing or security testing. The audit report says which applies.

From audit to testers in your sprint

  1. Functional test audit: about one week. A written report, requirements mapped against your tests, and a plan for the pilot.
  2. Pilot: two to six weeks. One product area with test cases, automated checks in your CI and a handover session.
  3. Ongoing: testers in your sprint and a written result after every release, as project outsourcing, team expansion or a dedicated team. See QA outsourcing, dedicated QA team and managed QA.

Each step ends with your decision on the next one. Prices are not published; describe the product and the quote comes within one business day. Every platform and test type we cover is listed under test automation 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.

  1. Step 1 About one week

    Functional test audit

    We read your stories, acceptance criteria and test cases side by side, and find the features nobody checks. Access to your tracker, environments and test data is agreed before we start.

    You get

    • Requirements and stories mapped against existing tests
    • Untested features ranked by business risk
    • Each existing automated test marked keep, rewrite or delete
    • What to automate and what stays manual
    • A report and a pilot plan written so your own testers can use them
  2. Step 2 Two to six weeks

    Pilot

    One product area. Test cases are written from its stories, the stable checks are automated in your CI, and testers run the rest.

    You get

    • Test cases in your tracker or test management tool
    • Automated tests in your repository
    • A handover session for your team
  3. Step 3 Monthly

    Ongoing

    Testers work inside your sprint, reading stories before they are built, and the same team moves repeated manual checks to scripts. You can buy it as QA outsourcing, as testers added to your team, or as a dedicated QA team.

    You get

    • A written result after every release
    • Monthly report
    • New features covered as they ship

From a test automation client

Zappletech have provided us with software development and test automation services, both in terms of guiding the architecture, and implementing a lot of test case.

Chris McVittie Head of Technology, UK More reviews

FAQ

Questions buyers ask

What do functional testing services include?

Checking features against their requirements. That means feature testing in the sprint, smoke checks on each build, system testing of full business flows before a release, and support for acceptance checks with your business users. Part of it is automated tests in your repository and CI, and part is testers, depending on what repeats. Re-checking existing features after a change is regression testing, and checking how services work together is integration testing; both have their own pages.

What is the difference between functional testing and system testing?

As we use the terms, functional testing checks a feature against its requirements, one feature or a group of features at a time. System testing runs the whole product, through full business flows, before a release. We do both, and the audit says how much of each your product needs.

What is not functional testing?

Checks of how the product behaves rather than what it does, such as speed under load or resistance to attack. They have their own pages, performance testing and security testing. We also treat checking how services and third-party systems work together as its own kind of work, integration testing.

Is functional testing manual or automated?

Both. Checks that repeat on stable features are automated, in your repository and CI. New and changing features, and anything that needs judgement, are checked by testers first, and repeated manual checks are moved to scripts by the same team once the feature settles.

How do you write test cases from our requirements?

Testers read your stories and acceptance criteria while they are written, and turn each criterion into a test case. Where a requirement is unclear or missing, we raise it as a question to you instead of guessing. AI can draft test ideas and test cases from the stories; the tester corrects every draft before you see it.

What is functional testing as a service?

The same work bought as an engagement instead of hired one person at a time. It can run as project outsourcing, team expansion or a dedicated team, or as managed QA where we own the result against measures agreed at the start. See managed QA.

Do you help with user acceptance testing?

We support acceptance checks that you run with your own business users, in a separate environment with production-like data. We help prepare the checks and the data. The decision to accept the release stays with you. For more on the practice itself, see our article on user acceptance testing.

Which platforms do you cover?

Functional testing applies across platforms. For the automation side, see web test automation, mobile test automation, API testing and desktop application testing, each with its own page.

How much do functional testing services cost?

Prices are not published. What moves the quote is the number of features, how often you release, how much is automated and which platforms are involved. Describe the product and you get a scoped quote within one business day.

Start with a one-week functional test audit

Tell us what your product does, how your requirements are written today, and what you test with, if anything. We reply within one business day with scope, timeline and a quote.

  • Founded by the founders of CodeceptJS and Codeception, open-source testing frameworks
  • Founded by the founders of Testomat.io and Testomat.ai
  • 32 test automation engineers and experts
  • 42 completed test automation and infrastructure projects
  • On the market since 2015
  • Rated 4.9 out of 5 on Clutch, from 7 client reviews

Prefer to talk first? Schedule a call

    We reply within one business day.