AI makes claims. I test them.
Know what your tests and AI agents are actually proving.
I help engineering leaders improve unreliable automation, evaluate AI systems and make release decisions using evidence rather than confidence.
Microsoft MVP, Developer Technologies
LinkedIn Learning instructor, 90,000+ learners across three courses
Former BBC Principal Test Engineer
20+ years in software engineering
Consulting
Practical help for teams that need better evidence.
Three bounded engagements. Each has a defined scope, defined deliverables and a defined ask of your time, so you know what you're buying before we start.
Diagnostic · two weeks, part-time
AI Quality & Engineering Productivity Diagnostic
from £4,500fixed price, agreed before we start
Where is delivery friction coming from, and is your AI adoption helping or hiding it? An evidence-based assessment with a prioritised improvement plan.
What's included
- Interviews
- Up to 8 stakeholder interviews (engineering, QA, product, delivery), 45 minutes each
- Scope
- One product team or up to 3 repositories; larger scopes quoted separately
- You provide
- Read access to repositories and CI, recent pipeline and defect data, current test strategy documents, access to interviewees
- Deliverables
- Written assessment across delivery flow, test strategy, automation effectiveness and AI tool usage; a scored maturity view; a prioritised 90-day improvement plan; a 60-minute readout to leadership
- Your time
- About 8 hours across the two weeks, plus the readout
- Not included
- Implementation of the plan, tool procurement, hands-on code changes (see the Sprint)
For: Heads of Engineering and QA who have been asked "why is delivery slow?" and want an answer backed by evidence.
Enquire about the Diagnostic
Pilot sprint · five working days
Playwright Modernisation Sprint
from £7,500fixed price, agreed before we start
A focused pilot that gives your team a migration strategy, reference implementation and measurable path to a faster, more reliable test suite.
What's included
- Scope
- One repository or test suite; migration from Selenium, Cypress or legacy Playwright patterns
- You provide
- Repository and CI access, a representative slice of the suite (typically 10 to 20 tests), one engineer as counterpart
- Deliverables
- Migration strategy document; framework architecture (fixtures, structure, configuration); reference implementation of the agreed slice in your repository; CI configuration for parallel runs; a measured before-and-after on the slice; two enablement sessions
- Your time
- Counterpart engineer available about 2 hours a day; two 90-minute team sessions
- Not included
- Migrating the whole suite (quoted after the pilot), ongoing maintenance, infrastructure changes outside CI configuration
For: Teams losing time to slow feedback, flaky results and automation they cannot confidently trust.
Enquire about the Sprint
Workshop · one day, up to 12 people
AI Agent Testing Workshop
from £4,500fixed price, remote or on site
Before an agent acts on your users' behalf, learn how to test what it actually did, and what evidence to keep.
What's included
- Content
- Agent behaviour, tool use and failure modes; evaluation methods; LLM-as-judge limitations, evaluator bias and appropriate use; Playwright-based validation of agent actions with retained evidence
- Format
- Hands-on, using a prepared example agent and test project so every attendee runs the checks themselves
- Your system
- Applying the checklist to your own product is optional and subject to access, suitability and preparation agreed in advance
- Deliverables
- Workshop materials, the example project, the AI Agent Testing Checklist, and a 30-minute follow-up call within two weeks
- Your time
- One day for attendees; one 30-minute prep call with the organiser
- Not included
- Building or fixing your agent, ongoing evaluation work (see the Diagnostic)
For: Teams shipping agents, and the engineers being asked to sign them off.
Enquire about the Workshop
Not sure which fits? Request a discovery call. Remote, or on site in the UK and Europe. Prices exclude VAT.
Proof
Credentials you can check, and work you can measure.
- Microsoft MVP, Developer Technologies. MVP profile
- LinkedIn Learning instructor, 90,000+ learners across three courses covering Playwright and Selenium. Instructor page
- Former BBC Principal Test Engineer, responsible for testing strategy on national-scale services.
- 20+ years in software engineering, spanning development, test automation, quality engineering and engineering leadership.
- Speaker at BCS SIGiST and SEETEST 2026 on AI testing.
Case study · anonymised, shared with permission
Playwright modernisation, [sector] product team
- The problem
- [e.g. A Selenium suite of N tests took X minutes in CI and failed intermittently, so the team re-ran builds and stopped trusting results.]
- My responsibility
- [e.g. Technical lead for test; owned the migration strategy and reference implementation.]
- What I changed
- [e.g. Moved the suite to Playwright with fixtures and project-level parallelism; removed duplicated coverage; introduced trace-on-failure.]
- The measured result
- Suite runtime from [baseline] to [result]; flaky failure rate from [before] to [after].
- How it was measured
- [e.g. Median CI duration over N runs before and after, same agent pool; flake rate as retries per 100 runs.]
Courses
Build faster, more reliable test automation.
Three LinkedIn Learning courses covering Playwright and Selenium, with 90,000+ learners between them.
[ Official course cover: Playwright Design Patterns ]
LinkedIn Learning · Playwright
Playwright Design Patterns
Fixtures, page objects and project structure for suites that keep growing. 71,000+ learners.
View course →
[ Official course cover: Advanced Playwright Techniques ]
LinkedIn Learning · Playwright
Advanced Playwright Techniques
Optimising speed, stability and cloud testing: parallelism, sharding and tracing. 14,000+ learners. Updated April 2026.
View course →
[ Official course cover: Learning Selenium ]
LinkedIn Learning · Selenium
Learning Selenium
Structure, scale, run and optimise automated tests. 4,500+ learners.
View course →
Free videos on Playwright and AI testing: YouTube · LinkedIn
Free download
The AI Agent Testing Checklist
27 checks before you trust an agent in production: boundaries, tool-use verification, the check step, failure modes, evaluation, and the Playwright evidence to keep for each. Three pages. The same checklist used in the workshop.
Speaking
Talks on AI testing, Playwright at scale and the human side of quality.
Conference talks and workshops for engineering audiences. Practical, evidence-led, and honest about what AI tools can and can't do yet.
Speaking enquiry
Appearances
- The Human Side of AI TestingBCS SIGiST · 7 July 2026
- Testing AI Agents with Playwright and MCP: A Practitioner's GuideSEETEST 2026 · upcoming
About
Uncle Aaroh is Qambar Raza.
Engineering Productivity & AI Quality Leader at CGI, Microsoft MVP and LinkedIn Learning instructor. Formerly Principal Test Engineer at the BBC, with 20+ years in software engineering spanning development, test automation, quality engineering and engineering leadership.
Based in Manchester, UK. Views are my own. Not affiliated with Microsoft's Playwright team.
My work sits where software engineering, quality and AI meet. I focus on a simple question: what evidence do we have that the system behaves as expected, especially when automation or AI gives us a confident answer?
