QA Pipeline Audit
$1,900 fixed
1 week
A clear picture of where your testing stands, and what to fix first.
- Review of your tests, CI and release process
- Risks ranked by impact
- A written plan with effort estimates
- A walkthrough call with your team
Agentic QA for software teams
Every feature ships with tests. Automatically.
I build QA into your development lifecycle with AI agents. Every feature and every bug fix comes with tests, and your team keeps shipping at full speed.
14+ years in QA automation for Cisco, Adobe, Yara, LottieFiles and others.
The problem
Fast teams outgrow their testing
Startups and scale-ups ship fast. Testing is the first thing to slip, and the last thing anyone has time to fix.
No one owns quality
You have 5 to 50 developers and no QA. Users find the bugs before you do.
No one trusts the tests
The suite is slow and flaky. People rerun it until it passes, or stop running it.
Coverage falls behind
Features ship faster than anyone writes tests for them. Every release is a bet.
How it works
Five steps from “we should test more” to tests on every change
Audit
I review your code, pipeline and existing tests, and find where bugs get through.
Plan
We agree on what to cover first, based on risk and how your team ships.
Build the foundation
I set up a test framework that fits your stack, with clear conventions, and run it in CI on every pull request.
Add the agents
I set up AI agents with rules for your repo. Every feature and bug fix gets tests, written with the code.
Hand over
Your team owns it. I document how it works, and stay on if you want ongoing help.
Services and pricing
Start small, or go all in
Clear scope, clear price. Not sure where to begin? Start with the audit.
Core service
Agentic QA Setup
from $5,500
2–4 weeks
A test framework and AI agents, built into how your team already works.
- Test framework for UI and API, in your repo
- Tests run in CI on every pull request
- Agent rules so new code arrives with tests
- Flaky tests in your current suite fixed or removed
- Documentation and handover to your team
Ongoing QA Partner
from $1,200 /month
Monthly
I stay on as your part-time QA engineer.
- Maintain and extend the test suite
- Review test coverage on new work
- Tune the agents as your code changes
- A monthly note on quality and risk
Prices in US dollars.
Case study · Enterprise SaaS knowledge platform
From 100 smoke tests to about 5,000 automated tests
- automated tests, up from 100
- ~5,000
- client configurations covered
- 1,000+
- adaptive UAT tests
- 400
- for ~50 tests, down from a month
- 2–3 days
Framework refactor
I rebuilt the test framework. The suite grew from 100 smoke tests to about 5,000 automated tests, and it has stayed stable since.
UAT for 1,000+ client configurations
Every client had its own configuration. I built a UAT framework with a binding layer and a test reference JSON: 400 adaptive tests that adjust to each client. Built in four months. Zero per-client maintenance.
Migration validation
The platform moved from a legacy VM system to a new Kubernetes deployment. I built tests that compared the two, so the team could switch with evidence, not hope.
Parallel agents
With AI agents working in parallel, output went from about 50 tests a month to about 50 tests in 2–3 days.
Built this way
I run my own products on the same system
What I set up for clients, I use every day. Here is the proof.
Domaćini
A directory that helps people in Serbia find local food producers and contact them in one tap. Built and run by me.
- 1,194 automated tests: 966 unit and integration (Vitest), 228 end-to-end (Playwright)
- End-to-end suite runs in 4 browsers: Chromium, Firefox, WebKit and mobile Chrome
- Every pull request: build, type check and unit tests in GitHub Actions, then the full Playwright suite on an isolated staging copy of the app
- Built with Claude Code: project rules, 40 skill files, and every test traced from feature to test case to coverage report
Stack: Next.js, Fastify, PostgreSQL, Playwright, Vitest
Visit oddomacina.rs (opens in a new tab)More coming soon
The next project is in progress. It will be built and tested the same way.
About
You work with me, not an agency
I'm Nikša Klecak. I've spent 14+ years in QA automation: UI, API and mobile, for Cisco, Adobe, Yara, LottieFiles and others.
Since 2021 I've been a technical screener for the Toptal network.
I'm based in Belgrade and work remotely with teams in the US and Europe. QALITY is a one-person service: you work with me directly, from the first call to the handover.
My LinkedIn profile (opens in a new tab)FAQ
Questions teams ask first
What does “agentic QA” mean?
AI agents write and maintain tests as part of your normal workflow. I give them rules and conventions for your repo, so the tests they write are consistent and easy to review. People stay in control: every test goes through code review like any other change.
Do you replace a QA team?
No. If you have no QA, I set up the system and stay on as long as you need. If you have QA engineers, I help them cover more with less manual work.
Which tools do you use?
Mostly Playwright and TypeScript for UI and API tests, Claude Code for the agents, and the CI you already have, such as GitHub Actions or GitLab CI. I adapt to your stack rather than replace it.
Who owns the code?
You do. Everything I build lives in your repository, and your team can run it without me.
What do you need from us?
Access to the code and CI, and one person on your team who can answer questions about the product.
How do we start?
Email me a few lines about your team and your testing. We'll set up a 30-minute call. If it's a fit, we start with the audit.
Contact
Tell me where testing hurts
Send a few lines about your team, your stack and your current tests. We'll set up a 30-minute call to see if I can help.
Prefer LinkedIn? Message me there (opens in a new tab).