Speaking

Talks and workshops on release decisions, quality gates and QA work when AI writes the code.

I am Michał (Mike) Buczko, Software Quality Competence Lead at Carrier/Viessmann. I approve product software releases across 25+ R&D teams and 400+ engineers, after nearly 20 years in software quality. On the side I built BowSmith, an iOS app, from idea to App Store in 20 months by directing AI agents. My talks answer one question: how do you know, with evidence, that software is good enough to ship, and what changes when AI writes the code.

Book a 30-minute call · Sessionize profile · LinkedIn

Talks I give now

All six are available in English and Polish. I also join panels and run internal company sessions.

Who Signs the Release When AI Writes the Code?

Format: Leadership talk, 40–45 minutes
For: Quality leaders, engineering managers, CTOs and CIOs

AI assistants write a growing share of production code. Somebody still has to say "this is good enough to ship", and in most organisations nobody has redefined what that signature should rest on. I connect two views: approving releases across 25+ R&D teams, and 20 months of shipping an app with AI agents.

You will leave with:

• The evidence to demand before a release when the author is a machine
• Where to place quality gates in an AI-assisted pipeline, and which ones to drop
• A way to position QA people as operators and judges of AI output

Don't Cut Your Testers, Promote Them: Quality Engineers as Your Missing AI Operators

Format: Leadership talk or keynote, 30–45 minutes
For: CIOs, VPs of Engineering, HR and transformation leads

When AI writes the code, the first saving many organisations reach for is the QA team. I argue it is the most expensive saving a CIO can make: the people best placed to judge AI output are the ones who already judge software for a living. I show what the AI operator role is and how to move quality engineers into it.

The Speed Trap: Reporting AI Productivity to the Board Without Fooling Anyone

Format: Leadership talk, 30–45 minutes
For: CIOs, IT leaders who report to a CFO or board, PMO and delivery heads

AI made code generation an order of magnitude faster, and verification did not keep up. Most AI productivity numbers count what was generated, not what the organisation can defend. I show how to tell the two apart and how to report AI-assisted delivery honestly.

From Tester to AI Operator: 20 Months of Shipping with AI Agents

Format: Experience talk, 30–45 minutes
For: All levels

The story of the BowSmith build, started with zero iOS knowledge. I started with 32 role-based agents; only 17 survived, and today six to eight remain, built around skills. I talk to one AI coordinator, which does the work itself or through subagents. I show what worked, what broke, and why testing skills are the core skills of an AI operator.

You will leave with:

• What broke with role-based agents, and the coordinator-and-subagents setup I use now
• Three testing habits an AI operator needs: ask what could go wrong, decide what evidence is enough, do not trust unchecked output
• A plain account of 20 months from idea to App Store

From Tester to AI Operator: Building the Career Path Before You Need It

Format: Talk, 40 minutes
For: Testers, QA leads and people leaders

Layoff data shows executional work going first while judgement holds. I turn that pattern, and what I learned shipping an app with AI agents, into a career path: the tester skills an AI operator needs, the ones to add, and how leads can plan the transition.

Design Your Quality Gates for AI-Assisted Delivery

Format: Half-day workshop, 3–4 hours, up to 24 people
For: Anyone who can bring one real delivery pipeline to work on

You bring one pipeline. You map where AI touches it and redesign each gate: evidence, reviewer, failure path. Then you test your design against failure cases from a real AI-built product.

You will leave with:

• A one-page gate map for your pipeline
• A release-evidence checklist
• Gate designs already tested against real failure cases

Watch

Critical Thinking Rules in AI-Enhanced Software Testing (Istanbul Software Testing Conference 2025, 26 min)

Framework for Thriving Remote Leadership (HUSTEF 2025, Budapest, 28 min)

In conversation with Richard Seidl (Software Testing Unleashed podcast, April 2026)

More recordings: Team Leadership in the Time of Remote Silence (PNSQC 2024, 48 min) https://www.youtube.com/watch?v=xEZMpHFRnwc

Upcoming

Sii x ŁuczniczQA meetup, Bydgoszcz, 25 Nov 2026. "Od testera do operatora AI" (From Tester to AI Operator), in Polish. · Talk page

No 2027 events confirmed yet.

Where I have spoken

• CAST 2026, Cocoa Beach, Florida, 3–5 Aug 2026. Invited speaker, 90-minute session "From Idea to Working Product: How AI-DLC Unlocks Human Creativity". Recap · Talk page
• Testing United 2025, Milan, 13–14 Nov 2025. "Is AI the future of Software Delivery: Can it replace a Feature team?" Recap · Talk page
• HUSTEF 2025, Budapest, 15 Oct 2025. "Building Trust, Fostering Growth, and Advocating Success for Remote Teams". Recap · Talk page
• QA or the Highway 2025, Columbus, Ohio, 27 Jun 2025. "Leading from Anywhere: Value-Based Framework for Remote Teams". Recap · Talk page
• test:fest 2025, Wrocław, 24 May 2025. "Critical Thinking rules in the AI-Enhanced Software Testing". Recap · Talk page
• Istanbul Software Testing Conference 2025, Istanbul, 22 May 2025. "Critical Thinking rules in the AI-Enhanced Software Testing". Recap · Talk page
• PNSQC 2024, Portland, Oregon, 14–16 Oct 2024. Talk and technical paper "Team Leadership in the time of remote silence". Recap · Talk page
• QA&TEST Embedded 2024, Bilbao, 9–11 Oct 2024. Talk on dashboards and KPIs. Recap
• QA or the Highway 2024, Columbus, Ohio, 21 Jun 2024. "Team Leadership in the time of being remote". Recap · Talk page
• test:fest 2023, Wrocław, Mar 2023. "YOU, the most important brand. Practitioner retrospective".
• SEETEST 2022, Sofia, 29–30 Sep 2022. "7 ways how to step up as a leader". Recap
• 2015–2019: test:fest (eight editions as a speaker in total), HUSTEF 2018 ("Automated Security Scanning in Payment Card Industry"), TestCon Europe, Testing United, TestWarez, SEETEST 2019 and others. A decade of test:fest

Book

I wrote the chapter "Quality at Machine Speed" in "The AI Transformation", a multi-author book published in 2026. It is available on Amazon.

For organisers

Everything a programme committee usually asks for. Copy what you need.

Short bio (about 50 words)

Michał (Mike) Buczko is Software Quality Competence Lead at Carrier/Viessmann, where he approves product software releases across 25+ R&D teams and 400+ engineers. He has nearly 20 years in software quality. He built BowSmith, an iOS app, by directing AI agents, and wrote a chapter in "The AI Transformation" (2026).

Long bio (about 120 words)

Michał (Mike) Buczko is Software Quality Competence Lead at Carrier/Viessmann. He approves product software releases across 25+ R&D teams and 400+ engineers, covering IoT, cloud, mobile, embedded and ASPICE-governed automotive software. He has nearly 20 years in software quality.

His work centres on one question: how do you know, with evidence, that software is good enough to ship, and what changes when AI writes the code. On the side he built BowSmith, an iOS app, from idea to App Store over 20 months by directing AI agents, starting with zero iOS knowledge.

Mike is the author of the chapter "Quality at Machine Speed" in "The AI Transformation" (2026). He has spoken at CAST, PNSQC, HUSTEF, Testing United, SEETEST and test:fest. He is based in Poland and speaks in English and Polish.

Speaker details

• Name form: Michał (Mike) Buczko
• Languages: English and Polish
• Travel base: Wrocław, Poland
• Formats: Keynote or talk (30, 45, 60 or 90 minutes), half-day workshop, panel, internal company session
• Tech needs: Just a projector to connect to.
• Workshop room: table layout for 5-6 small groups, flipcharts, projector
• Sessionize: sessionize.com/michal-buczko
• More about me: About page

Book me

The fastest way is a free 30-minute call. Tell me the event, the audience and the date, and we will pick the talk that fits.

Book a 30-minute call

Prefer to write first? Message me on LinkedIn. Programme committees can pull my details from Sessionize.