Shubham Shende

About

By the Numbers

QA Engineer who builds test systems that actually prevent production failures.

I started out running manual test cycles on LMS and ERP modules, and taught myself automation because I was tired of catching the same regressions by hand. Today I'm a QA Analyst at Studium Tech, where I own the end-to-end automation strategy across web UI and REST API layers in fast-paced Agile sprints.

My focus is on writing tests that genuinely prevent production failures - not just inflate coverage numbers. Every test targets a real failure mode, and I lean on AI-assisted QA techniques to sharpen test generation and defect analysis.

0+ Test Cases Delivered Across web UI & REST API layers

0% Regression Reduction

0+ Years in QA

0 Critical Production Defects

Stack

Technical Skills

Automation frameworks, programming languages, API testing, and AI-assisted QA.

Automation

Playwright Selenium WebDriver Pytest Page Object Model Data-Driven Testing Cross-Browser Testing End-to-End Testing

API & Database

REST API Testing Postman JSON Schema Validation Auth Flow Testing HTTP Status Verification SQL MongoDB Assertions

Programming & Tooling

Python Git GitHub Jira VS Code Chrome DevTools AI-Assisted Authoring

Testing Practices

Functional & Regression Smoke & Sanity Integration Testing Black Box Testing UAT SDLC · STLC Agile · Scrum Defect Lifecycle

Projects

Selected Work

Test frameworks that ship. API suites that catch. Pipelines that never sleep.

Authentication Regression Suite

regression-suite-login

Problem: Login, invalid-credential and forgot-password paths - the highest-risk flows - were re-tested manually on every build.
My Role: Authored the suite and its page objects.
Approach: Selenium + Pytest on a Page Object Model with a shared BasePage; positive, negative and boundary auth scenarios, centralised config, screenshot capture and an Excel run report.
Outcome: Repeatable authentication regression run; reduced repetitive manual checking of login paths.

Selenium WebDriver Python Pytest Page Object Model openpyxl

Issue Triage & Tracking Platform

issue-tracker-app

Problem: Product issues raised by the Customer Support team had no structured path into engineering, so reports arrived incomplete and unreproducible.
My Role: QA owner for triage - I reproduce each reported issue, judge severity, and coordinate the fix with the development team.
Approach: Validation across the platform's REST API surface (auth, issues, users, clients, notifications) and its role-based access rules, backed by MongoDB checks on the stored records.
Outcome: A defined Support → QA → Engineering path: issues arrive reproducible, triaged and verified on fix.

REST API Testing RBAC Validation MongoDB Defect Triage Bug Reproduction

Featured Project

Event Management E2E Framework

github.com/shubhamshendeworks/playwright-app-framework

Problem: Regression across the event-management platform - login, dashboard, event creation, editing and categorisation - was re-checked by hand every release: slow, repetitive, and inconsistent between runs.
My Role: Designed and built the framework end to end - architecture, page objects, fixtures, test-data strategy, utilities and reporting.
Approach: A layered Playwright + Pytest framework on the Page Object Model, with Excel-driven test data, reusable utilities, centralised configuration, saved authentication state to avoid repeated logins, and automatic screenshot capture on failure. AI-assisted tooling (Claude Code / MCP) was used locally to accelerate script authoring.
Outcome: Repetitive manual regression replaced by a repeatable suite; maintainability improved through POM; reusability increased via modular utilities; execution and reporting standardised (HTML report, run log, Excel run summary).

Playwright Python Pytest Page Object Model openpyxl pytest-html

View on GitHub View README

Framework Architecture

Test Layer Pytest specs - login, core flow, event creation, edit & category

Page Layer · POM Login · Dashboard · Events List · Event Form · Event Details · Category

Fixture Layer conftest.py - browser setup / teardown, saved auth state

Data Layer Excel-driven test data via openpyxl

Utility Layer Helpers · Logger · Network · Screenshots · Report manager

Config & Reporting Central settings · HTML report · Run log · Failure screenshots

pages/page objects - one class per screen tests/test modules + conftest fixtures test_data/externalised test inputs utils/logger · screenshots · reporting · helpers configs/environment & browser settings reports/HTML report · run log · run summary

How I Work

Engineering Approach

How a requirement becomes a verified release - the path I take on every feature.

01 Requirement Review the spec and acceptance criteria before code exists.
02 Test Case Design Functional, negative and boundary cases mapped to the requirement.
03 Manual Validation Explore the flow by hand first - understand it before automating it.
04 API Validation Verify auth, contracts, status codes and stored data beneath the UI.
05 Automation Design Promote the stable path into a page-object-backed script.
06 Regression Suite Fold it into the suite so it is re-run on every release, not by hand.
07 Bug Verification Reproduce, log, re-test on fix, and close only when verified.

Framework Design Decisions

Page Object Model

One class per screen. Locators and actions live behind the page object, so a UI change is a one-file fix - not a suite-wide rewrite.

Test Data Management

Inputs are externalised into Excel and read at run time, so the same script covers many data sets without edits.

Utilities & Helpers

Shared modules for logging, network handling, screenshots and reporting - written once, reused across every test.

Configuration Management

Environment and browser settings are centralised, so switching target or browser never means touching a test.

Fixtures & Auth State

Setup and teardown live in fixtures; a saved authentication state skips repeated logins and keeps runs fast and isolated.

Reporting & Screenshots

Every run produces an HTML report and a log; failures auto-capture a screenshot named by test ID, reason and timestamp - so a red run is diagnosable without re-running it.

Principles

01

Automate the Repetitive

Spend human attention on judgement, not on clicking.

02

Maintainability Over Speed

Maintainable frameworks over quick fixes.

03

Shift Left

Test the requirement, not just the build.

04

Two Lenses

Think like an end user, then like an attacker.

05

Coverage With Intent

Never a number to game.

06

Reusable Architecture

Over one-off scripts, every time.

07

Self-Explaining Failures

A failing test must explain itself.

Experience

Timeline of roles, responsibilities, and impact delivered.

Studium Tech Pvt. Ltd. Software Test Engineer Quality Assurance Engineer QA Analyst Jul 2023 – Present

Cut test authoring effort by 40% - Introduced MCP-based AI script generation into the QA pipeline across 3+ automation projects.
Own the automation strategy - Scalable E2E suites in Playwright, Selenium and Pytest, applying Page Object Model and data-driven patterns across UI regression and smoke workflows.
15+ API endpoints under test - Token auth flows, JSON schema contracts, HTTP responses and error handling verified via Postman, plus role-based access scenarios across 3+ user types.
Keep the pipeline green - Own GitHub Actions test health: resolve flaky patterns, enforce zero-defect release gates, and drive risk assessments, sign-offs and defect triage each sprint.

Tech Stack

Playwright Selenium WebDriver Python Pytest MCP Automation GitHub Actions Postman REST API Jira

Cut manual execution by 30% - Kicked off and scaled the first Selenium WebDriver + Python regression suite, covering 20+ scenarios within the first 3 months.
Zero critical defects at release - Ran Functional, Regression, Integration, Smoke and Sanity cycles across 10+ product modules every sprint.
Set the API testing standard - Repeatable Postman patterns for 10+ endpoints (auth, JSON schema, HTTP responses), backed by MongoDB assertions that caught field-level mismatches before they reached the application layer.
Traceable coverage, tracked defects - Maintained a spec-mapped test case library and drove Jira defects from reproduction through to verified closure.

Tech Stack

Selenium WebDriver Python Pytest Postman MongoDB REST API Jira SQL

Owned QA for LMS & ERP modules - Functional, smoke, sanity and regression cycles verifying core workflows, UI behaviour and business-rule compliance.
Cross-browser & mobile coverage - Validated layout consistency and navigation across Chrome, Firefox, Edge, Safari and Android, surfacing platform-specific rendering and interaction issues.
Designed the test case library - 100+ structured cases built from business requirements, covering functional, negative and boundary conditions with clear pass/fail criteria.
80+ defects raised and closed - Detailed Jira reports with reproduction steps, environment, severity and screenshots - plus UAT coordination and post-release spot checks.

Tech Stack

Manual Testing Cross-Browser Testing Black Box Testing Jira UAT Test Case Design

Education