
Sign up to save your podcasts
Or


On this episode of Into the MoTaverse, Rosie Sherry talks to Sam Webb, lead test engineer at the Driver and Vehicle Standards Agency (DVSA). After years of relying on vendors and contractors, the DVSA is bringing software engineering in house, and Sam has been leading with the growth of his team. The conversation picks up from Leading with Quality in Brighton, the day before MoTaCon 2026.
How quality engineering creates room for depthSam explains why he split one overloaded role into test engineers and software development engineers in test (SDETs), who build test frameworks and tooling for the whole delivery team. The aim is depth: test engineers stop repeating checks and focus on risk-based testing and quality decisions. He also covers faster feedback loops across test levels, and shift right, which means monitoring software in production to check it does what the business needs.
Building buy-in and quality cultureSam is candid about the culture change, from his own team’s concerns to developers looking at him “as if I had six heads”. He shares why a head of quality has to be part salesperson, using analogies and Jira data to win buy-in from developers, product owners and BAs.
Follow Sam Webb.
Follow Rosie Sherry.
Episode highlights:From e-learning to the DVSA: Sam’s 18 years in testing, from building the Welsh government’s e-learning platform to brand protection and counterfeit detection.
Bringing engineering in house: Why the DVSA is moving away from outsourced vendors and what that means for quality.
In-house versus contractor quality: How delivery pressure on contractors can push automation and quality aside.
Retiring the ninja unicorn: Splitting one overloaded role into test engineers and SDETs, and avoiding automation versus manual silos.
Defining quality engineering: Testing, quality control, quality assurance and quality engineering, plus the quality role of product owners, BAs and UCD.
Who owns quality?: Shared story points, developers contributing to tests, and why the head of quality sometimes has to be a salesperson.
Metrics as storytelling: The Jira data showing work spiking on days four, seven, nine and ten of the sprint, and why you should embrace the bug.
In the bin: Test efficiency, test levels, shift right, and why knowing good quality puts testers in pole position for AI.
Explore more about quality engineering in the MoTaverseWant to keep going? These are good next steps from the MoTaverse community:
New to quality engineering? Start with The Community’s Guide to Quality Engineering, a collection of articles, talks and definitions from quality people.
Curious about the SDET role? Read the SDET definition in the MoT software testing glossary.
Want to balance shift left with shift right? Explore The Community’s Guide to Shifting in Testing: Shift Left, Shift Right, Shift Everywhere, then watch We can test in production: An introduction to shifting testing right.
Interested in where quality engineering is heading? Read The community writing trend towards quality engineering, continuous quality and holistic testing.
Listen to more conversations like this one in the [Into the MoTaverse testing to quality engineering series].
🏆 Ready to build your quality engineering skills? Get qualified with the MoT Software Quality Engineering Certificate.
Louise Woodhams is automation and test lead at ZapMap, the app that maps EV chargers across the country, and a TWiQ leader in the MoTaverse. She joins Rosie to talk about joy at work when tech feels relentless. Louise shares the small things that keep her going: Wicker Man quotes in her Cypress test data, a piggy bank icon in the internal admin panel, a Gemini-generated seal of approval on every release, and a corgi spotted mid-spiral. They also revisit the day a critical PHP security fix had her team watching GitHub, Bitbucket and Down Detector until 8pm, and why that counts as trauma bonding.
The conversation gets candid about AI. Louise explains why she and her testers have chosen not to use it, how developers are now worried about their jobs too, and what gets lost when a favourite dev starts talking to Claude instead of the tester. Rosie shares why the MoTaverse is doubling down on humans, what owning her own content has taught her, and how Claire Norman’s MoTaCon talk on glimmers shaped both their thinking. Plus: mermaids are safe until 2096.
Follow Louise Woodhams
Follow Rosie Sherry
Is your team busy fixing symptoms while the real problems sit somewhere else in the system? Melissa Fisher, quality and test consultant and MoTaCon workshop host, joins Rosie Sherry to talk about quality management systems (QMS) and what software teams can learn from them.
Melissa explains QMS through three areas: quality planning, quality control and continuous improvement. She defines quality as the absence of friction and shares a simple exercise she ran at her MoTaCon workshop: map what your team actually does in each area, look for where things stick, and repeat every six to twelve months as AI changes how work gets done. She also covers how low, medium and high product risk should shape your system, and why QMS, born in manufacturing and ISO 9001, deserves a place in software.
Rosie adds her take on systems thinking, quality orchestration and The Ahh Test principle behind Team MoTaverse decisions. A practical listen for anyone who wants to improve the system, not just test what it produces.
Follow Melissa
Follow Rosie
"How do you sleep at night?" is a question the MoTaverse team asks internally when making decisions, with a specific focus on the AI world,, so Rosie put it to two people who have spent decades building tools for teams who cannot afford a bad night. Arthur Hicken has been at Parasoft since 1992. Igor Kirilenko joined fourteen years ago and now leads the product.
Arthur's answer starts with the 2am phone call. You can stay in firefighting mode forever, or you can stop lighting fires, which means thinking ahead about the crazy user, the cyber attack, the slow network, the ten million users you never planned for. Igor's answer is visibility. Do you know what you tested, can you repeat it, and can you even finish the suite overnight?
Then AI arrives and unsettles both answers. They cover why the "let AI discover and test everything" pitch has gone quiet, the repeatability problem that killed it, test impact analysis for rerunning only what a build touches, service virtualization for non-deterministic dependencies, automated remediation, and why safety critical teams carry the most worry but sleep the easiest.
Follow Arthur Hicken and Igor Kirilenko and Rosie Sherry
Joe Colantonio runs TestGuild. It has been twelve years since Rosie was last on his podcast, and long enough for both of them to have watched the same ideas come round more than once.
The loop is hard to miss. Testing skill mattered, then tooling mattered, then everyone had to become a developer, and now AI writes the code and testing skill matters again. WinRunner would fix reliability, then Selenium would, then Playwright. Keyword-driven testing and behaviour-driven development were both going to let everyone test. Joe watched both roll out at GE and neither stuck. He started an AI testing summit in 2018 to near silence.
And yet they are both clearly enjoying themselves. Rosie rebuilt her personal site with Lovable, imported her old tweets, and built Spaceducking as a deliberately buggy practice site. Joe vibe coded a test tool matcher and an accessibility scanner, neither of which he needed. Half the time goes on testing what the AI produced, which is the point.
Resources:
The Ahh Test
Quality is Care
TestGuild
Dr Michaela Greiler has spent over a decade researching code review, first at Microsoft Research and the Visual Studio team, then as Head of Research at DX, and now through her own research and consultancy work. In this episode she talks to Rosie Sherry about what agentic development is doing to one of software's most central practices.
Michaela names two failure modes she's seeing in teams: code review exploitation, where verification work gets shifted onto a peer reviewer who never had the mental model in the first place, and code review surrender, where reviewers simply can't keep up with the volume and things get waved through. She walks through SCOPE, the model she's developing, which stands for Staged Code Oversight with Proportional Escalation, and explains why she thinks code review needs to start in the planning phase rather than at the pull request.
Michaela makes the case that reviewing test code, long the lowest priority on most teams, now matters more than ever, because agent-generated tests can signal safety where there is none.
Karim Jouini is CEO and co-founder of Thunders, an AI testing company built after he and his co-founder Jihed Othmani sold their expense management business, Expensya. He's never held a tester job title, but he's managed QA teams for most of his career, first at Microsoft and then across a fintech operating in sixty countries where regulation changed two or three times a week.
In this episode, Karim tells Rosie about beta testing the Xbox Kinect, a project he says was far harder to test than to build, and about the hiring culture he pushed back on at Microsoft, where strong engineers were steered into development and average ones into testing. He and Rosie land on "quality architect" as a title worth making popular, and talk through what that role actually needs: management skills, systems thinking, an understanding of the whole delivery loop, and product-oriented quality metrics.
Simon Tomes returns to Into the MoTaverse for a six-month check-in with Rosie on Community as a Service, care, joy, and the slow pivot to becoming a tech community.
Simon Tomes, the MoTaverse's Community Lead, returns for his second appearance on Into the MoTaverse, roughly seven months after his first, for a reflective check-in on what's changed.
Rosie and Simon revisit "Community as a Service," the idea that grew out of a fuzzy December/January conversation and has since become the backbone of decisions like making professional membership mandatory to contribute, introducing "layers of care," and building out the star and badge system.
They talk through the ongoing Chapters migration, the rise of Moments as a blogging platform baked into the star system, and the growing role of recorded conversations: Into the MoTaverse, Call for Insights, This Week in Quality, and round tables as a form of community service in themselves.
The conversation also covers the MoTaverse's long, deliberate pivot from a testing and quality community toward a broader tech community, including the new Engineering Managers chapter and the Leading with AI event run in partnership with PwC. Throughout, joy comes up again and again as the metric that matters most.
Follow Simon Tomes and Rosie Sherry in the MoTaverse.
Stephen Platten is a principal consultant at Inspire Testing, a MoTaverse ambassador, and the writer behind LinkedIn's "The Stoic Tester." A former Royal Air Force engineer who spent over a decade on fast jets before moving into trials, evaluation, and validation work, Stephen went on to lead test delivery at Yorkshire Water, work as a chapter lead at Admiral Insurance, and now consults across quality and testing for Inspire Testing.
In this episode, Stephen and Rosie explore how Stoic philosophy shaped his approach to critical thinking, and why that skill matters more than ever in the age of AI. They dig into Anthony Weston's Rule Book of Arguments, unpack logical fallacies like appeal to automation, appeal to authority, false dilemma, hasty generalisation, and survivorship bias, and discuss how testers can use bottom-up, systems thinking to challenge AI-generated content.
The conversation also covers root cause analysis and human factors borrowed from Stephen's Air Force background, why AI strategy gets more attention than AI governance, and how evidence-based arguments, not stronger opinions, should win the day.
Follow Stephen Platten and Rosie Sherry in the MoTaverse.
Adriana Vilela is a Principal Developer Advocate at Dynatrace and a core contributor to OpenTelemetry, the open source standard for instrumenting applications that has become one of the highest-velocity projects in the Cloud Native Computing Foundation.
In this episode, Adriana and Rosie trace the arc of long careers in tech, from Java development in the early 2000s to the current age of AI-assisted coding, and reflect on what has actually changed and what hasn't.
They dig into what observability means for testers (more than most people think), the idea of telemetry as a quality gate, and why OpenTelemetry's vendor-neutral community model is a rare success story worth studying.
The conversation also navigates into AI slop, the flood of low-effort AI-generated pull requests threatening open source communities, and what project maintainers, community builders, and organisations can actually do about it.
If you work in quality, DevOps, or community, this one connects threads you probably haven't seen joined up before.
From the publisher's feed