QA Expertise for AI-Powered B2B Sales Platform

Challenge

• A dedicated QA owner to guarantee release quality for a live product with 150+ paying customers
• Order in the bug backlog, where outdated and already-fixed issues were mixed with real ones
• Systematic test coverage of business-critical flows: lead creation, routing, enrichment, and CRM data synchronization
• Reduction of technical debt in the existing e2e test suite, which was unstable in CI

Solution

• Audited the entire bug backlog: identified stale and already-fixed issues, produced cleanup reports, brought the board to an actionable state
• Created a dedicated QA board, test plans for new features, and a regular bug-triage process
• Established a clear verification loop: every fix was retested before release
• Dashboard: account tables, filters, reports, permissions, data tables redesign, getting-started flows
• Knock Card on web, iOS, and Android — onboarding, saved-card flow, Apple Wallet and Google Wallet integration, design conformance, mobile resolutions
• Routing engine: meeting and chat routing, routing rules, Routing Simulators, rep availability, CRM-owner-based routing
• Sales tools: scheduling links (default and private), personal links, Sales Rep Hub, Marketing Card and its analytics
• Communication features: chat links, chat transcription, monitoring messages, FAQ publishing, Telegram and Slack apps, LinkedIn invite connections, Zoom integration
• Web Tag, Website Intent, and Website Conversion tracking — including backend verification
• Deep testing of HubSpot, Salesforce, and Marketo integrations: enrichment flows, mapped fields, CRM prioritization, account owner logic, new lead creation, CRM backfill, and CRM data monitoring
• Built a Playwright proof-of-concept, then finalized and merged the autotests PR with stabilized CI runs
• Revived and fixed skipped autotests (Administration, Routing Rules modules)
• Created the e2e test codebase for the knock-connect application from scratch
• Set up CI rules for test runs and explored AI-assisted testing tools, including writing reusable Claude skills to speed up QA workflows

Result

Predictable, low-risk releases, a transparent QA process, reliable CRM data flows, and an automation suite that catches regressions before customers do.