Recap: 7th Berlin QA Meetup – Lessons from Trade Republic, JetBrains, and InDrive

Vitaly Sharovatov -

The meetups just keep getting better!

This time, despite the perfect weather, we had an incredible turnout — with familiar faces returning for their fifth or sixth time and even some attendees traveling in from other German cities. I’m beyond thrilled to see our community growing and staying so engaged.

Here’s a recap of the talks that sparked vibrant (and sometimes heated — in the best way!) discussions:

🚀 1. Vasiliy Bolgar – How Trade Republic Achieves 500 Deploys a Day Without QA Engineers

Vasiliy gave us a deep dive into how Trade Republic ships over 500 deployments a day — all without a traditional QA team.

Their philosophy: quality is a platform capability. It’s baked into the system through:

  • Autotests
  • CI/CD pipelines
  • Automated rollbacks
  • Observability
  • Feature flags

The core insight? The biggest blocker to safe, frequent releases isn’t bugs — it’s deployment anxiety. By building confidence into the deployment process, developers are empowered to ship continuously and fearlessly.

Testing in production is the norm. They use targeted rollouts, real-time monitoring, and shift focus from bug count to Mean Time to Recovery (MTTR) as the key quality metric. The role of QA evolves — from bug-hunting to enabling speed, safety, and resilience through tools and system design.

🔥 This talk generated tons of passionate debate — and we loved every minute of it.

🧪 2. Denis Mashutin – Smarter Regression Testing at JetBrains (Without More Tools or People)

Denis, from the PyCharm QA team at JetBrains, showed us how a simple idea led to big results.

He introduced “shuffle checks” — randomized subsets of their manual regression test plan, spread across early releases. No new tools. No extra headcount.

What it changed:

  • Broke routine and reduced bias
  • Caught critical issues earlier
  • Encouraged testing of unfamiliar areas
  • Reduced pressure in late stages
  • Freed up time for deeper QA work

It was a great reminder that powerful QA improvements can start with nothing more than a spreadsheet, a script, and a team open to trying something new.

🚕 3. Arsenyi Batyrov – Dogfooding at Scale: Lessons from a Ride-Hailing Giant (InDrive)

Arsenyi, Head of Manual QA at InDrive, delivered one of those rare talks that offered something for everyone: testers, managers, and change-makers alike.

He shared how InDrive transformed dogfooding into a scalable QA engine, surfacing real issues across extreme diversity — 49 countries, hundreds of device models, various payment providers, and unpredictable network conditions.

But this wasn’t just “try using the product.”

It was a systemic QA process — complete with real-time reporting, automated classification, Jira integration, and dashboards — all run without developers. The small QA team used ChatGPT and Python to automate:

  • Team assignment
  • Bug categorization
  • Issue transitions
  • Dashboarding

This allowed them to build and iterate without waiting on engineering, making QA self-sufficient.

Arsenyi also challenged conventional wisdom:

“Metrics are only useful when they reveal problems.
Field testing isn’t about edge cases — it’s about understanding your users’ real environments.”

More than a technical case study, this was a powerful story of organizational feedback loops, QA empowerment, and bridging the gap between users and builders.

Thanks again to all the speakers and attendees. If this momentum continues, our next meetup might need a bigger room!

Want to join the next one? Stay tuned — we’d love to have you.