QA Automation Engineer | UI, API, Database and AI-assisted Testing
I focus on building test solutions that give teams useful feedback—not just more test scripts. My work covers modern browser automation, API validation, database testing, CI/CD and the manual QA practices that support confident releases.
I care about readable tests, stable pipelines, clear failure evidence and coverage that reflects real product risk.
| Area | Tools and practices |
|---|---|
| UI automation | Playwright, Cypress, Selenium WebDriver, Page Object Model |
| Languages | TypeScript, Java, Python, SQL |
| API testing | REST Assured, HTTP validation, contract and negative testing |
| BDD and test runners | Cucumber, Gherkin, TestNG, JUnit |
| Data quality | PostgreSQL, ETL reconciliation, source-to-target validation |
| CI/CD | GitHub Actions, cross-browser execution, reports and artifacts |
| Manual QA | Test planning, exploratory testing, test cases, defect reporting |
| AI-assisted QA | GitHub Copilot, MCP workflows, Langfuse and LLM evaluation concepts |
| Project | What it demonstrates |
|---|---|
| Playwright UI Automation | TypeScript end-to-end framework with reusable fixtures, Page Objects, cross-browser execution, traces and CI |
| Selenium Java Automation | Java, Selenium and Cucumber framework with explicit waits, reusable components and GitHub Actions |
| Playwright Cucumber | Business-readable Gherkin scenarios connected to a maintainable Playwright implementation |
| REST Assured API Automation | Positive, negative and schema-focused API coverage with reusable request specifications |
| Cypress Automation | UI interactions, network interception, API contract checks and visible cross-browser results |
| SQL Data Validation | PostgreSQL checks for completeness, relationships, financial reconciliation and ETL accuracy |
| Software Test Cases | Risk-based manual cases, smoke and regression suites, requirements traceability and execution reporting |
| Software Bug Reports | Reproducible defect reports with impact, severity, priority, evidence guidance and retest criteria |
- Start with customer and business risk before choosing what to automate.
- Keep test intent visible and move repeated technical details into reusable components.
- Use API and database checks where the UI alone cannot prove the result.
- Separate smoke feedback from broader regression coverage.
- Treat screenshots, traces, logs and test counts as part of the framework—not an afterthought.
- Write documentation so another tester or developer can continue the work without a handover meeting.
I am developing practical ways to use AI in testing while keeping human judgment and verification at the centre of the process.
My current focus includes:
- using GitHub Copilot to accelerate test scaffolding, refactoring and documentation while reviewing every change
- connecting tools and test context through Model Context Protocol (MCP)
- using Langfuse concepts to trace prompts, model responses, latency, cost and failures
- designing repeatable evaluations for LLM response quality and regression
- testing AI features for unsafe input, inconsistent output and observability gaps
I see AI as a way to improve coverage and investigation speed—not as a replacement for thoughtful test design.
I am continuing to deepen my work in Playwright, API automation, SQL validation, CI reliability and AI-enabled testing workflows. I am particularly interested in teams where QA contributes early to requirements, architecture and release decisions.