Headline

Lead Product Quality Engineer @ Curately AI | HR Tech — AI recruiting, Chrome extension, ATS integrations | Python · Playwright · pytest | Building agentic QA tooling (MCP) | LLM evaluation | BFSI background

About

I test non-deterministic systems for a living, and I build the tooling that makes it repeatable. Over the past years I've owned end-to-end quality across two regulated, integration-heavy domains — BFSI lending platforms, and now HR Tech. Page Object Model frameworks, UI and API automation suites, and regression coverage that holds up under frequent releases. Across the last fifteen months that has meant 591 commits, 95 pull requests authored, 155 reviewed and 544 delivered tasks across 18 repos and products. Python and Playwright are my current stack; Java and Selenium are where I built the foundation. The tooling changes, the work doesn't — find where a system breaks before a user does. At Curately AI I lead quality for an AI recruiting platform: the web application, a Chrome extension, ATS integrations across vendor systems, and the backend event consumers behind them. ATS integration is where HR Tech gets hard — every vendor models candidates, jobs and applications differently, and a sync that silently drops a field is a defect nobody notices until a recruiter does. The other half is newer. A growing share of what I test is non-deterministic, and conventional assertions don't survive contact with it. A regression suite built on exact output equality is the wrong instrument for a model that paraphrases. Working out what the right instrument looks like is the problem I find most interesting right now. My answer so far has been to build the tooling rather than wait for it. Over the last quarter I built an agent-driven QA layer on top of our stack: a ticket-to-test-plan-to-execution chain, browser skills that drive the web app and the Chrome extension including unpacked builds loaded over CDP, two-tab flows that create a record in an external ATS and verify the sync landed in-product, and ten MCP integrations spanning code search, logs, product analytics and databases — with a SELECT-only guardrail so no automated session can mutate QA or production data. Roughly 900 database checks and 150 log queries have run through it. Reliability engineering for agents turns out to be the same discipline as test design, applied one level up. From here I'm going deeper into LLM evaluation and AI system testing — eval design, LLM-as-judge scoring, and observability — applied to a real codebase rather than to exercises. Open to remote roles where quality engineering is treated as core to delivery, particularly on teams shipping AI products. Python · Playwright · pytest · Selenium · Java · SQL · API testing · MCP & agent tooling · CI/CD

Location

Hyderabad

AI stack

antigravity

claude code

claude

cursor

Devin

Wispr Flow

Work experience

Lead Product Quality Engineer at Curately AI, Inc, Mar 2026 - Present · 7 mos, • Own end-to-end quality for an AI recruiting platform in HR Tech — web application, Chrome extension, ATS integrations and the backend event consumers behind them — as DRI for ATS integrations and post-sales engineer for client implementations. • Built an agent-driven QA workflow covering test-plan generation, live execution and write-back reporting, which now handles the majority of ticket verification and collapses the path from ticket to written-up result into a single command. • Integrated Several MCP servers across code intelligence, work tracking, browser automation, log aggregation, product analytics and QA/production databases; wrote a SELECT-only enforcement hook so no automated session can mutate data — database checks have run under it. • Verify candidate, job and application sync across multiple third-party ATS vendors, each with its own data model and webhook behaviour — building two-tab flows that create a record in the external system and confirm it landed correctly in-product. • Built browser automation that drives the web application and loads unpacked and store-published Chrome extension builds over the Chrome DevTools Protocol. • Ran release validation for four consecutive Chrome extension releases, and root-caused a filter regression between two builds by diffing tags, commits and pull requests across repositories. • Repaired and extended the Playwright/Python suite flow by flow — exploring each flow live to capture real selectors before fixing its page objects, instead of batch-patching brittle locators. • Built a read-only seven-phase discovery harness for onboarding a new ATS integration, capturing entity models, API surface, webhook behaviour and field mappings, and ran it end to end against a new vendor. • Mentor teammates on test design, automation practice and exploratory testing technique across a fully remote team.

Senior Product Test Engineer at Curately AI, Inc, Aug 2025 - Feb 2026 · 7 mos, • Stood up the first automated coverage for the AI sourcing web module and the Chrome extension's sourcing flows, taking both from manual-only to repeatable suites. • Designed and executed functional, regression and smoke test strategies across web workflows and backend services, establishing the release-gate checks the team still runs. • Built and maintained UI and API automation in Python and Playwright, introducing Page Object structure so suites stayed maintainable as the product changed weekly. • Found and drove closure of a production data-exposure defect in which contact details were returned in an API response without the gating user action having been performed. • Reported production defects with reproducible evidence across search filters, result counts and AI-match behaviour, several of which traced back to endpoint wiring rather than UI logic. • Partnered with product managers and engineers to refine acceptance criteria before development started, moving quality input earlier in the cycle. • Analysed production defects for recurring patterns and fed them back into regression coverage to reduce escapes over successive releases.

Senior QA Automation Engineer at Virinchi Ltd., May 2022 - Jul 2025 · 3 yrs 3 mos, Hybrid, • Owned functional and regression automation for BFSI applications — loan origination and servicing platforms — building test plans, scripts, scenarios and test data in Java and Selenium WebDriver. • Designed and maintained a Page Object Model automation framework with Extent Reports integration, improving reusability and cutting the time to diagnose a failing run from the report alone. • Automated end-to-end loan workflows including application, approval and payment processing across high-volume transactional systems. • Implemented backend and API automation with REST Assured, and SQL-based database validation to confirm transaction accuracy. • Executed smoke, functional, GUI, regression and cross-browser compatibility testing via Selenium Grid across supported environments. • Continuously refactored test assets — identifying coverage gaps, refining scenarios and updating automation to keep pace with frequent product change in a regulated domain.

Education

Bachelor's degree, Commerce, Osmania University