QA Automation Architect with 15+ years in enterprise quality engineering across web, microservices, and mobile app development in regulated financial domains.
If you review only five repositories, review these:
- selenium-e2e-automation-framework: Selenium WebDriver 4 + Java 17 + JUnit 5 + Allure: 15 UI journeys on Swag Labs, 5 REST Assured API tests, with all 20 tests on Jenkins.
- enterprise-qa-automation-blueprint: Enterprise Playwright framework: BDD, DB validation, Pact contract tests, k6 performance, ZAP security, with 419 tests running in CI.
- appium-e2e-automation-framework: Appium + Java 17 + JUnit 5: 15 native mobile tests on the Swag Labs app, Android emulator + iOS simulator, Allure reports, with 15 Android tests.
- karate-bdd-api-testing: Karate + Java 17 + JUnit 5: BDD API tests against the ReqRes API, 11 scenarios across 4 feature files.
- cypress-e2e-automation-framework: Cypress + JavaScript: Page Objects, fixtures, custom commands, UI + API specs, mochawesome and Allure reports.
Read in this order, they cover five areas: a proven automation framework, a full enterprise QA blueprint, native mobile automation, BDD API testing, and Cypress E2E automation.
Beyond traditional test automation, this work covers AI systems, evaluation harnesses, and test generation with local models. Everything runs locally: no cloud model is called, and no test run leaves your machine:
- sidekick-rag-assistant: Enterprise RAG assistant (LangChain + FAISS): pluggable document sources, cited answers, React chat UI, MCP server, retrieval evals. No API keys needed.
- crewai-scrum-team: A multi-agent Scrum team (PO/BA/SM/Developer/SDET) on CrewAI + local Ollama that plans, builds, and validates software.
- llm-eval-harness: 50 golden Q/A items, heuristic + LLM judges, drift detection (PSI), model comparison dashboard. It answers which model to ship, with evidence.
- doc-extract-eval: Compares 5 PDF extraction engines (pypdf, pdfplumber, PyMuPDF, Tesseract, Mistral OCR) with ground-truth F1 scoring, then sends the winning extraction through an LLM for structured business summaries.
- ollama-junit-demo: Local Ollama drafts JUnit 5 unit tests for a small Java app. Drafts live in ai-generated/ until a human reviews and promotes them. Coverage via JaCoCo, execution reports via Allure.
- ollama-jev-decision-demo: Typed decision models via Ollama (/v1/systemone): choice, yes/no, and score questions with calibrated probabilities.
| Skill from profile | Repository | What to check | Proof artifact |
|---|---|---|---|
| E2E web automation frameworks | selenium-e2e-automation-framework | Selenium 4 Page Objects, 15 UI journeys + 5 API tests, Jenkins CI | All 20 tests passing on Jenkins + Allure report |
| Native mobile test automation | appium-e2e-automation-framework | Appium Page Objects, 15 tests on Android emulator + iOS simulator | All 15 Android tests passing; iOS 9 passed, 6 intentionally skipped |
| Enterprise QA strategy and test architecture | enterprise-qa-automation-blueprint | Playwright BDD, DB validation, Pact contract tests, k6 performance, ZAP security | 419 tests passing in CI |
| Cypress E2E automation | cypress-e2e-automation-framework | Page Objects, fixtures, custom commands, UI + API specs | mochawesome-report/ + allure-results/ per run (CI template in ci/) |
| Contract testing and integration risk control | microservices-contract-and-integration-testing | Multi-service orchestration scenarios: happy-path, timeout-fallback, schema-drift. Pact consumer + provider verification, OpenAPI validation, WireMock virtualization | docs/evidence-suite.md |
| Release governance and CI/CD quality gates | cicd-quality-gates-reference-implementation | PASS, WARN, BLOCK decisions from metrics plus integration report input | docs/evidence-integration.md |
| Multi-agent AI engineering | crewai-scrum-team | Five-agent Scrum team on local Ollama: planning gates, ATDD acceptance tests, validation against the original spec | outputs/validation_report.md |
| Enterprise RAG systems | sidekick-rag-assistant | Cited answers, permission-aware retrieval, hybrid mode, retrieval evals | evals/ regression suite |
| LLM evaluation and model selection | llm-eval-harness | Golden dataset, judge agreement, drift detection, side-by-side model comparison | reports/ |
| Document AI evaluation | doc-extract-eval | Extractor comparison with ground-truth F1 scoring | reports/ |
| AI-generated unit tests (local LLM) | ollama-junit-demo | Ollama drafts JUnit 5 tests for a Java app, human review before promotion | ai-generated/ |
- Quality architecture: end-to-end automation strategy, CI/CD pipelines, quality gates, and release decisions grounded in test evidence
- API and integration testing: contract testing (Pact), service virtualization (WireMock), integration validation, REST Assured
- AI-assisted quality: evaluating LLM systems with evidence: golden datasets, LLM-as-judge scoring, drift detection across runs; AI-drafted Playwright and JUnit tests from local models, promoted only after human review
- Platforms and tools: Playwright, Selenium, Cucumber, TypeScript, Java, Python; performance testing (JMeter); mobile automation (Appium, Sauce Labs, LambdaTest)
- Domain and team: quality engineering across banking, financial services, credit lending/payments, and mortgage-backed securities; hands-on mentoring of junior SDETs in automation practices and code quality
- LinkedIn: https://www.linkedin.com/in/acbharath14/