AI & Innovation
What We're Building
Teststack Labs is developing a proprietary AI-assisted quality engineering framework — designed to make automation smarter, faster, and more resilient. This is our R&D frontier, not a product for sale.
The Problem We're Solving
Modern software delivery moves faster than traditional quality engineering can keep up with. Test suites become brittle. Coverage gaps widen. Engineers spend more time maintaining automation than building it. AI changes that equation — but only if it is applied with engineering rigour, not hype.
We are building tooling that augments the QE engineer — not replaces them. The goal is a framework that learns from your codebase, adapts to change, and surfaces risk intelligently. Built by QE engineers, for QE engineers.
Capability areas
What We're Building
AI-Assisted Test Generation
Automatically generate meaningful test cases from source code, API contracts, and user stories. The system analyses change impact and proposes coverage — engineers review, refine, and approve.
- AST-based code analysis for test surface mapping
- OpenAPI / GraphQL schema ingestion
- Natural language story-to-test translation
- Coverage gap identification
- Engineer-in-the-loop approval workflow
Self-Healing Automation
Locator drift is one of the biggest causes of flaky tests. Our self-healing layer detects broken selectors at runtime and proposes repairs — reducing maintenance overhead without sacrificing reliability.
- Multi-strategy locator fallback
- DOM similarity scoring
- Automated repair suggestion with confidence scoring
- Audit trail for all healing events
- Playwright-native integration
Intelligent Risk Scoring
Not all code changes carry equal risk. Our risk scoring model analyses commit history, code complexity, and test coverage to prioritise which tests to run — and which areas need more attention.
- Change impact analysis from VCS history
- Cyclomatic complexity weighting
- Historical defect density mapping
- Dynamic test suite prioritisation
- CI/CD pipeline integration
Anomaly Detection in Test Results
Pattern recognition across test run history to surface anomalies — flaky tests, performance regressions, and coverage drift — before they become production incidents.
- Time-series analysis of test run data
- Flakiness scoring and trend detection
- Performance baseline deviation alerts
- Coverage regression identification
- Slack and pipeline notification hooks
Natural Language Test Authoring
Enabling QA engineers and business analysts to describe test scenarios in plain language and have executable automation generated — with full transparency into what was produced and why.
- LLM-powered intent parsing
- Step-to-code translation with review layer
- BDD Gherkin output support
- Human review and override at every step
- Framework-agnostic output targets
QE Observability Dashboard
A unified view of quality signals across your pipeline — test health, coverage trends, defect escape rates, and automation ROI — designed to give QE leads and engineering managers the data they need to make decisions.
- Cross-pipeline test result aggregation
- Coverage and quality trend visualisation
- Defect escape rate tracking
- Automation ROI metrics
- Exportable reports for stakeholders
How We're Building It
Engineer-in-the-loop
AI surfaces suggestions. Engineers make decisions. We are building augmentation tools, not autonomous systems.
Transparency first
Every AI-generated output is explainable. No black boxes. Engineers can see why a suggestion was made and override it.
Framework-agnostic
Built to integrate with the tools teams already use — Playwright, Cypress, k6, GitHub Actions, Jenkins, and beyond.
QE-native design
Designed by QE engineers who live in the problem space — not AI researchers building tools they have never used in production.
Where We Are
Self-healing POC
Proof-of-concept for self-healing locator repair validated in controlled test environments.
Test generation & risk scoring
Active development of AI-assisted test generation and intelligent risk scoring modules.
Observability & anomaly detection
QE observability dashboard and anomaly detection layer — planned for development following Phase 2 validation.
Integrated framework
Unified AI-assisted QE framework combining all modules — target for limited early-access programme.
Interested in what we're building?
We are selectively working with engineering teams who want to be part of the early-access programme. If that sounds like you, let's talk.