Hi, I'mNavin Kumar — SDET
Building AI-Augmented Test Frameworks for enterprise-scale software, blending traditional automation rigor with modern LLM & agentic engineering.
Quality engineering, powered by AI
Navin Kumar
I'm a Senior SDET with 11+ years of experience in Java-based test automation, object-oriented design, and AI-driven quality engineering across Banking, Payments, Insurance, and Telecom domains. My work centers on building test systems that don't just verify software — they think alongside it.
Over the last few years I've gone deep into LLM evaluation, RAG pipeline validation, prompt engineering, and agentic workflow automation — including building Cursor AI–driven pipelines that delivered 3x productivity gains for QA teams. I've tested vector embedding retrieval accuracy, semantic evaluation metrics, and LLM hallucination detection using DeepEval, and I hold an AWS Generative AI certification.
Equally at home in classic automation — Python, Playwright, Selenium, Docker, Kubernetes, Jenkins, GitHub Actions, REST/SOAP API testing — I bridge the gap between traditional QA discipline and the emerging world of MCP-based agentic orchestration and multi-agent workflow design.
Tools I build quality with
A blend of core automation engineering and next-generation AI/GenAI testing capability.
Automation & Languages
- Java
- Python
- OOP
- Selenium WebDriver
- Playwright
- Appium
API & Data
- REST Assured
- SOAP UI
- Postman
- SQL
- Database Testing
AI / GenAI Testing
- DeepEval
- Prompt Engineering
- RAG Validation
- GitHub Copilot
- Cursor AI
- AWS GenAI
DevOps & Tools
- Docker
- Kubernetes
- Helm Charts
- Jenkins
- GitHub Actions
- Git / Bitbucket
- Maven
- JIRA / HP ALM
Where I've built quality at scale
- Maintained a Java-based REST Assured framework validating 35+ network API endpoints, integrated into Jenkins CI/CD for continuous quality on every sprint.
- Designed an agentic QA pipeline using Cursor AI that reads Jira tickets & HLDs to autonomously generate test cases and automation scripts — cutting authoring time in half and delivering 3x productivity.
- Built a Playwright-based AI skill executing full end-to-end UI regression flows without human intervention.
- Validated RAG-based LLM pipelines on AWS using DeepEval — testing vector retrieval accuracy and reducing hallucination rates in production GenAI features.
- Led Java-based API & UI automation across 15 Franklin Templeton investment products, sustaining near-zero production defect escapes for 3+ years.
- Drove defect triage and root cause analysis, reducing defect escape rate to under 1% via test governance and release-gate enforcement.
- Mentored 4 junior QA engineers through code reviews, standards, and framework onboarding.
- Led mobile QA for the OLA Foods Android app, improving test coverage by 95% with zero showstopper defects at launch.
- Automated 30+ payment & order management APIs with REST Assured, saving 36 hours of manual regression per sprint via Mountebank mock servers.
- Built Appium-based automated mobile tests, boosting product reliability.
- Automated API & microservice testing using Selenium WebDriver with Java and Python.
- Extended Selenium suites each sprint, collaborating with dev teams on defect resolution.
- Built REST Assured & SOAP UI frameworks from scratch, automating 300+ manual banking API tests.
- Cut regression execution from 5 days to 1 day per release — an 80% reduction that accelerated release throughput.
- Analyzed API/backend logs in Splunk for end-to-end issue resolution.
- Translated user stories into automated test scripts using UFT and Quality Center for onshore teams.
- Maintained scripts through monthly release cycles, ensuring ongoing test effectiveness.
Selected work & initiatives
Frameworks and pipelines I've designed and shipped across enterprise QA teams.
Agentic QA Pipeline
Cursor AI–driven pipeline that reads Jira tickets, acceptance criteria & HLDs to autonomously generate test cases and Python scripts, committing directly to Bitbucket — 3x overall productivity gain.
Autonomous UI Regression Skill
Playwright-based AI skill that reads test steps from Jira and executes full end-to-end UI flows without human intervention, eliminating manual regression effort every sprint.
RAG / LLM Evaluation Framework
Validated RAG-based LLM response pipelines on AWS — testing vector embedding retrieval accuracy and automated semantic evaluation metrics to reduce model hallucination.
Franklin Templeton Test Automation
Selenium BDD & SOAP UI frameworks across 15 investment products (mutual funds, SIP, portfolio management) sustaining near-zero production defect escapes for 3+ years.
Westpac Core Banking Automation
Built REST Assured & SOAP UI frameworks from scratch, automating 300+ manual banking API tests and cutting regression time from 5 days to 1.
OLA Foods Mobile QA
End-to-end mobile QA for an Android food-delivery app using Appium and Mountebank mock servers — 95% coverage improvement, zero showstopper defects at launch.
Foundation & continuous learning
B.Tech / B.E. — Information Technology
🏅 Certifications
- ✓ AWS Partner: Generative AI Essentials
- ✓ Generative AI Learning Plan — Technical Professionals
- ✓ Introduction to Prompt Engineering with GitHub Copilot
- ✓ Using GitHub Copilot with Python
- ✓ Introduction to GitHub Copilot
- ✓ Python Basics
Let's build something reliable, together.
Open to Senior SDET, QA Architect, and AI Test Engineering roles. Whether it's a framework audit or a full agentic QA build-out — my inbox is open.