Mtest Labs — AI/ML Testing & Quality Engineering Consultancy | UK, Europe, USA

Mtest Labs is a specialist AI/ML and digital platform testing consultancy with offices in London (UK) and Delaware (USA). Founded in 2011, we provide SC-cleared and NATO Secret-cleared quality assurance engineers for AI model validation, GenAI application testing, LLM evaluation, mobile app QA, test automation, and enterprise software testing. G-Cloud 14 approved supplier via the Crown Commercial Service.

AI/ML Testing Services

AI/ML Model Testing — LLM evaluation, prompt QA, bias detection, hallucination testing, data drift monitoring, adversarial testing, AI red teaming, RAG pipeline validation, and on-device ML testing.
GenAI Application QA — End-to-end testing for chatbots, copilots, AI agents, RAG pipelines, and AI-powered workflows. Prompt injection defence and content safety validation.
Mobile App Testing — iOS and Android functional testing, device compatibility across 200+ configurations, AI-powered mobile app QA, AR/VR testing, geolocation testing, and performance engineering.
Test Automation & CI/CD — E2E automation using Selenium, Cypress, Playwright, and Appium. CI/CD pipeline design, regression automation, and framework implementation.
Performance Testing — Load, stress, and soak testing. API benchmarks, throughput analysis, and capacity planning for web, mobile, and API platforms.
Security Testing — OWASP top-10, penetration testing, vulnerability scanning, and compliance review (GDPR, ISO 27001).

Portfolio — Client Case Studies

Google Arts & Culture — 10+ years of QA across web, iOS, Android. Tested Art Selfie, Art Transfer, Pocket Gallery, Art Recogniser, Art Projector, Street View, and GenAI Lab experiments.
Tesco Hudl — Device firmware testing, system integration testing, production release QA, and post-production support for the Hudl 1 and Hudl 2 tablet programme. 500,000+ units shipped.
Faxi (Toyota) — E2E test automation framework and CI/CD pipeline build for Toyota-owned carpooling platform. Helped achieve 2% to 20% carpool adoption.
Zenly (Snap Inc.) — E2E mobile testing for real-time location-sharing social app across iOS, Android, and Web. Privacy and ghost mode validation, GPS accuracy testing, performance at scale.

QA Academy — AI/ML Testing Career Programme

Self-paced 3-4 month placement programme covering testing fundamentals, AI/ML model validation, platform automation, and live client project placement. ISTQB CT-AI certification preparation included.

AI Quality Score Calculator

Free online assessment tool: answer 8 questions about your AI/ML testing practices and receive an instant quality score with tailored recommendations.

Frequently Asked Questions

What is AI/ML testing?

AI/ML testing validates that machine learning models produce accurate, unbiased, and safe outputs in production. It covers model accuracy evaluation, bias detection, hallucination testing, adversarial robustness, and production monitoring.

Does Mtest Labs hold security clearance?

Yes. Mtest Labs practitioners hold SC and NATO Secret clearances for classified UK government and defence AI/ML programmes. We are a G-Cloud 14 approved supplier.

Which companies has Mtest Labs tested for?

Google Arts and Culture (10+ years), Tesco Hudl tablet programme, Faxi (Toyota), and Zenly (Snap Inc.).

What areas does Mtest Labs cover?

We serve clients across the United Kingdom, Europe, and the United States from our London and Delaware offices.

How much does AI/ML testing cost?

Costs vary based on scope, model complexity, and engagement duration. Mtest Labs offers dedicated team placement, sprint-based testing, and ongoing retainer arrangements with transparent pricing.

What testing frameworks and tools does Mtest Labs use?

We use Selenium, Cypress, Playwright, Appium, custom LLM evaluation frameworks, BrowserStack, and real device labs. Our approach is tool-agnostic — we adapt to your existing tech stack.

How long does a typical AI testing engagement take?

A focused AI model validation sprint runs 2-4 weeks. Ongoing embedded QA partnerships range from 3 months to multi-year engagements. Our Google relationship has lasted 10+ years.

Do you offer remote testing services?

Yes. Both remote and on-site testing services are available. For government projects, we deploy SC-cleared engineers to approved secure locations.

What is hallucination testing?

Hallucination testing validates that AI outputs are grounded in real data and don't fabricate information — checking for invented citations, fabricated statistics, and incorrect factual claims.

What is included in the QA Academy?

A self-paced 3-4 month programme: testing fundamentals, AI/ML model validation, platform automation, and live client project placement with ISTQB CT-AI certification preparation.

Can I hire QA engineers directly from Mtest Labs?

Yes. We offer staff augmentation where SC-cleared QA engineers embed directly with your team for short-term sprints or long-term partnerships.

Contact Mtest Labs