// QA Lead · Automation Engineer · AI Testing Specialist

Biswajit
Satapathy

7+ Years AI-Driven QA Team Lead

QA Lead at VyaparApp leading a 15-member team across desktop, mobile & web platforms for India's #1 SMB software (15M+ users). Builder of AI-native Playwright frameworks using Claude & Gemini APIs — cutting test creation from hours to minutes.

↗ Get in Touch View Projects
Biswajit Satapathy
01 // impact

Numbers That Matter

15M+
Users on VyaparApp — the product I ensure quality for
80h→4h
Desktop regression runtime via AWS Spot parallel execution
~90%
Reduction in manual test writing via AI intent pipeline
2,600+
Automated test scenarios (Desktop 1,600 + Mobile 1,000)
30min
Test case creation — down from 4–5 hours with AI pipeline
15
QA engineers mentored & led across desktop, mobile & web
02 // skills

Core Competencies

🤖 AI & LLM Integration
Claude API Gemini API Prompt Engineering Agentic AI RAG LLM Test Generation Ollama
🧪 Test Frameworks
Playwright (TS) Selenium Appium Cucumber BDD TestNG REST Assured
💻 Languages
Java TypeScript JavaScript Python SQL HTML/CSS
☁️ CI/CD & Cloud
Jenkins GitLab CI AWS EC2 S3 IAM Spot Instances AMI
📋 QA Practices
Test Strategy Shift-Left Agile/Scrum SDLC/STLC Exploratory Testing DB Verification
🛠 Tools & Infra
Git Jira Confluence Postman MongoDB Self-healing Selectors
03 // experience

Career Timeline

QA Lead Current
VyaparApp — Vyapar Accounting & Invoicing
July 2024 – Present · Bengaluru, India
  • Lead a team of 15 QA engineers delivering quality across desktop (Windows/Mac), mobile (Android/iOS), and web platforms for 2–3 releases/month.
  • Architected an AI-native Playwright framework for Vyapar's Electron desktop app — the org's first AI-driven automation. Reduced test creation from 4–5 hours to 30 minutes.
  • Built Multi-Phase AI Agent Execution where Claude writes Playwright JS at runtime — handles scrolling, dropdowns, overlays autonomously. Self-heals YAML test files.
  • Serve as release stakeholder, coordinating dev, product, and support teams for release readiness.
  • Led end-to-end QA hiring drives to grow and strengthen the team.
Senior Quality Engineer
VyaparApp — Vyapar Accounting & Invoicing
Sep 2022 – Jun 2024
  • Built & maintained the desktop automation framework (Java + Selenium + WinAppDriver + Cucumber BDD) with 1,600+ test scenarios.
  • Implemented AWS Spot-based cloud execution pipeline via Jenkins — reduced regression from 80+ hours to 4 hours with significant cost savings.
  • Led mobile automation framework (Java + Appium + Cucumber BDD) with 1,000+ scenarios across multiple devices.
  • Improved mobile pass rate from 60% → 90–93% and desktop pass rate from 30% → 60% (95%+ in critical modules).
Quality Engineer 2
VyaparApp — Vyapar Accounting & Invoicing
Jan 2022 – Aug 2022
  • Executed REST API testing using REST Assured, Java, and Postman across multiple integrated modules.
  • Validated third-party integrations, test data workflows, and end-to-end data integrity.
Quality Engineer
VyaparApp — Vyapar Accounting & Invoicing
Nov 2018 – Dec 2021
  • Executed manual testing — functional, regression, exploratory, and smoke testing for 2–3 releases monthly.
  • Documented test plans and defect reports in Jira & Confluence; participated in Agile sprint ceremonies.
04 // projects

Key Projects

01
AI-Native Playwright Framework for Desktop Automation
End-to-end AI-powered framework for an Electron desktop app. Plain English YAML test cases executed by Playwright with AI element location via a four-phase engine: selector cache → direct Playwright → Claude AI (10 locator strategies) → AI Agent Mode (Claude writes unconstrained code at runtime). Self-healing YAML rewrites flaky steps automatically — getting cheaper with every successful run (~85% Phase 0 cache hit rate).
TypeScript Playwright Claude API Gemini 2.5 SQLite Node.js FFmpeg
02
AWS Spot + Jenkins Parallel Cloud Execution Pipeline
Jenkins pipeline that launches Windows EC2 Spot instances from a pre-configured AMI, fetches the Vyapar installer from S3, runs 1,600+ Cucumber scenarios in parallel, and tears down infrastructure on completion. Slashed desktop regression from 80+ hours to just 4 hours, with major cost savings via Spot pricing. Reports auto-published to Slack.
AWS EC2 Spot Instances Jenkins Java Selenium WinAppDriver Cucumber
03
Hackathon 2025 — Vyapar Knowledge Center (RAG Assistant)
On-premises, privacy-preserving knowledge assistant for contextual Q&A and guided desktop demos using full RAG architecture. Integrated PyAutoGUI for live desktop demonstration triggered by natural language queries — completely local with no data leaving the device.
Ollama ChromaDB SentenceTransformers Streamlit Python PyAutoGUI
04
Release Sanity Automation Tool
Internal Python tool for automated post-release sanity and version checks, eliminating repetitive manual validation steps in the release process. Packaged as an installable tool using NSIS for easy distribution to the QA team.
Python PyWinAuto OpenCV Selenium NSIS
05 // contact

Let's Connect

b7satapathy@gmail.com LinkedIn GitHub ↓ Download Resume