Functional & Exploratory Testing
Every user journey and business rule tested against written acceptance criteria, then off-script. Exploratory sessions target where bugs hide: sign-up and onboarding, checkout and billing, role-based permissions and data imports.
- Scripted and exploratory test sessions
- Acceptance testing against your user stories
- Reproducible bug reports with severity and video
Regression & Test Automation
End-to-end and regression suites that run on every pull request, in the framework that fits your stack, with tests your developers can read.
- Playwright, Cypress or Selenium for web UI
- pytest, Jest or Vitest for unit and integration
- Stable selectors, fixtures and seeded data
API Testing
Status codes, payload schemas, auth and permission boundaries, pagination, idempotency and error handling, checked against your OpenAPI spec.
- pytest with httpx, Postman and Newman
- Schema and contract checks (OpenAPI, Pact)
- Negative, boundary and auth-bypass cases
Performance & Load Testing
Load, stress, spike and soak tests that find where the system bends, with latency and error-rate thresholds as pass or fail checks.
- k6, JMeter or Locust scripts
- p95 and p99 latency, throughput, error budgets
- Bottlenecks across app, database and cache
Security Testing Basics
Baseline application security inside QA: OWASP Top 10 categories, broken access control, injection, auth and sessions, vulnerable dependencies.
- OWASP Top 10 checks, ASVS as the reference
- ZAP scans and dependency audits in CI
- Access-control tests across roles and tenants
Accessibility Testing (WCAG 2.2)
axe-core and Lighthouse paired with manual keyboard, screen reader, zoom and contrast checks, mapped to WCAG 2.2 success criteria.
- WCAG 2.2 Level AA audits
- NVDA, VoiceOver and keyboard-only testing
- Findings you can use in an ACR or VPAT
Mobile & Cross-Browser Testing
Real-device and browser coverage chosen from your analytics: iOS and Android apps, responsive web, and the versions your customers use.
- Appium for native and hybrid apps
- BrowserStack real-device and browser grid
- Responsive and visual regression checks
QA for AI and LLM Features
LLM features never answer the same way twice, so they need evaluation, not just assertions: eval suites rerun on every prompt or model change.
- Golden datasets from real user prompts
- Groundedness checks for RAG answers
- Prompt-injection and data-leak cases