$ whoami

Praveenkumar
Alagarsamy
Chandrasekaran

Lead AI Test Automation Engineer/Agentic AI Evaluation & LLM Assurance/Python & Test Harness Architecture

Open to work Fulda, Germany On-site · Hybrid · Remote

Portrait of Praveenkumar
PK · 2026
14.5+years in software quality
−45%manual testing effort with an AI test platform
+35%defect detection accuracy from AI coverage analysis
M.TechData Science & Engineering, BITS Pilani

01 / About

I build test harnesses for software that doesn't give the same answer twice.

I'm a Lead AI Test Automation Engineer with 14.5 years of software engineering experience, specialising in test harness architecture, non-deterministic system verification and Python-based automation framework development.

Right now I focus on AI assurance and evaluation: automated test suites that score LLM and agentic workflows on multi-turn state persistence, tool-calling accuracy, intent resolution and semantic assertions. I bring established software testing methods to new AI evaluation needs, turning system requirements and governance principles into reproducible, production-grade test pipelines.

Before AI, I spent over a decade on UI, API and performance automation for clients such as Amazon Audible, Overstock and athenahealth. I've led teams, mentored engineers and owned quality from requirements to release.

02 / Projects

Introlligent · 2023–2026

AI-powered test automation platform

A platform that tests AI features such as chatbots and AI suggestions. Prompting generates and runs Playwright checks, and Streamlit puts the results in front of the team.

−45% manual testing effort and a shorter test lifecycle

  • Playwright
  • Python
  • Streamlit
  • Prompt engineering

Introlligent · 2023–2026

LLM-as-a-Judge evaluation pipeline

A Python pipeline that uses a judge model to score LLM responses for semantic accuracy, hallucinations and overall response quality. It turns subjective review into repeatable, automated assertions.

Repeatable scoring for non-deterministic output

  • Python
  • LLM evaluation
  • Semantic assertions

Introlligent · 2023–2026

Open-weights model benchmarking

Benchmarked open-weights architectures against each other: Mixtral, a sparse Mixture-of-Experts model, and Jamba, a hybrid Mamba-Transformer. Measured quality and trade-offs to guide model selection.

Data-backed model choice

  • Mixtral
  • Jamba
  • Benchmarking
  • Python

Introlligent · 2023–2026

Evaluation dashboards & structured output

Streamlit dashboards that turn qualitative evaluation feedback into quantitative accuracy scores. Also designed JSON schemas and prompts that force LLMs to return structured output that can be tested.

Qualitative feedback becomes measurable

  • Streamlit
  • JSON Schema
  • Pandas

Introlligent · 2023–2026

AI-driven coverage & gap analysis

Brought AI into the QA lifecycle for test coverage and gap analysis, bug summaries and report generation, and led AI adoption across the testing team.

+35% defect detection accuracy

  • LLMs
  • Test coverage
  • Reporting

Introlligent · HR tech (Workday)

Workday UI & API automation

Led a team delivering Selenium UI functional automation, integrated with file-based and REST API automation in Karate, for HR-tech products built on Workday.

Better data accuracy, fewer manual errors

  • Selenium
  • Karate
  • REST APIs
  • Team lead

Relevantz · 2018–2023

Amazon Audible, Overstock & LoanBeam

Led UI and API automation teams across four major client projects, and ran JMeter performance and load testing to confirm scalability and find bottlenecks.

Met scalability benchmarks across 4 client projects

  • Selenium
  • Karate
  • Postman
  • JMeter

Relevantz · 2018–2023

Excel-based API testing tool

A custom in-house tool that validates backend API data from Excel-defined test cases, so testers no longer needed third-party software.

Faster API validation without licence costs

  • Java
  • REST
  • Excel

Relevantz · 2018–2023

Builder-pattern automation framework

Introduced the Builder design pattern into the automation frameworks, which made test data and page objects modular, reusable and stable. Recognised internally for code quality.

Recognised for code quality

  • Java
  • OOP
  • Design patterns

BORN Group · 2015–2018

Big Bus Tours & Tata CLiQ Luxury

End-to-end QA for omni-channel e-commerce across the US, UK and APAC on Magento, Hybris and Drupal. Worked as both Business Analyst and QA, from Statement of Work through to release.

BA + QA ownership, SOW to release

  • Magento
  • Hybris
  • Drupal
  • Agile

athenahealth · 2014–2015

Healthcare BI: Flash → HTML5 migration

Led testing of a US healthcare business intelligence tool through its move from Flash to HTML5, and built a Selenium hybrid automation framework from scratch for the repetitive checks.

Zero P0/P1 issues in UAT

  • Selenium
  • Hybrid framework
  • Regression

Amazon · 2010–2013

Kindle device testing & Sikuli POC

Cross-device testing and Build Verification Testing for Kindle apps, plus early automation POCs with QTP and image-recognition testing with Sikuli (computer vision).

More stable releases across devices

  • BVT
  • Sikuli
  • QTP
  • Devices

03 / Experience

  1. May 2023 – May 2026

    Lead Automation Engineer

    Introlligent Solutions · HR tech (Workday) · Bengaluru, hybrid

    • Built an AI-powered automation platform (Playwright, Python, Streamlit) that cut manual testing effort by 45%.
    • Designed LLM-as-a-Judge evaluation pipelines and benchmarked Mixtral against Jamba.
    • Led Selenium UI automation and Karate REST/file API automation, and pushed AI adoption across QA.
  2. Apr 2018 – Apr 2023

    Senior Software Engineer

    Relevantz Technology Services · Chennai

    • Managed UI and API automation teams (Selenium, Karate, Postman) for Amazon Audible, Overstock and LoanBeam.
    • Ran JMeter performance testing and built a custom Excel-based API testing tool.
    • Introduced the Builder design pattern into frameworks and was recognised for code quality.
  3. Oct 2015 – Apr 2018

    Senior Software Engineer

    BORN Group · E-commerce · Chennai

    • End-to-end QA for global e-commerce on Magento, Hybris and Drupal.
    • Worked as both Business Analyst and QA on Big Bus Tours and Tata CLiQ Luxury.
  4. Mar 2014 – Oct 2015

    Senior Software Engineer

    athenahealth · Healthcare · Chennai

    • Led testing for a healthcare BI tool through its Flash → HTML5 migration, with zero P0/P1 issues in UAT.
    • Built a Selenium hybrid framework from scratch.
  5. Dec 2010 – Mar 2013

    Device Associate

    Amazon Development Center · Kindle · Chennai

    • Cross-device Kindle app testing and Build Verification Testing, plus QTP and Sikuli automation POCs.

04 / Skills

Test automation

  • Selenium WebDriver
  • Playwright
  • Karate
  • Postman
  • JMeter
  • Protractor
  • Sikuli
  • QTP

AI & LLM evaluation

  • LLM-as-a-Judge
  • Agentic AI evaluation
  • Tool-calling accuracy
  • Prompt engineering
  • Structured output
  • Model benchmarking
  • Mixtral
  • Jamba

Programming & data

  • Python
  • Java
  • JavaScript
  • SQL
  • Pandas
  • NumPy
  • scikit-learn
  • Streamlit
  • REST / JSON

CI/CD & practices

  • Jenkins
  • Git
  • Agile / Scrum
  • TDD
  • OOP
  • Design patterns

Quality engineering

  • Test strategy
  • Estimation
  • Test design
  • Regression
  • BVT
  • ETL & DB validation
  • Bug triage

Leadership

  • Team lead
  • Mentoring
  • Business analysis
  • Stakeholder management
  • End-to-end QA ownership

05 / Education & languages

2021 – 2023

M.Tech, Data Science & Engineering

BITS Pilani, India

2005 – 2009

B.Tech, Information Technology

Sethu Institute of Technology, India

Languages

  • EnglishProfessional
  • GermanElementary · learning
  • FrenchElementary

06 / Contact

Let's talk about
testing what's hard to test.