Quality engineering for an AI-powered world.
AI can make software faster to produce. It does not make the hard decisions disappear: what good looks like, where failure matters, what evidence is enough, and who remains accountable. I bring 18 years across software quality, business analysis and technology delivery to those decisions, and now apply that foundation to the AI-enabled systems I build and evaluate.
Across automotive, insurance, financial services, medical and enterprise software, my work has repeatedly sat where business rules, requirements, risk and technology meet. That is where quality has the most leverage, and it is the same foundation I now bring to AI-enabled systems.
I bring structure to messy work, connect business problems with technical delivery, and look for ways to make the work easier to repeat, automate and improve. I am comfortable moving between understanding the problem and getting hands-on with the solution.
I organise requirements, decisions, data, workflows and learnings so people can see what matters, what is connected and what needs to happen next.
I can move between business, product, engineering and quality. I turn ideas and questions into something a team can build, test and make decisions around.
If a recurring problem can be solved with a small tool, automation or AI workflow, I am comfortable getting hands-on and making it real rather than stopping at the recommendation.
I document what we learn, remove repeated effort and keep refining the process. The goal is not just to solve today’s problem, but to leave behind something the team can keep using.
I care about the logic behind a decision and what happens when it meets reality. I dig until the problem makes sense, break it into something small enough to try, build or test it, and use what happens next to decide where to go.
For me, AI quality is not a check added at the end. It is making these questions answerable throughout the lifecycle.
Existing practice banks could be gamed through answer-position and option-length patterns, weakening the score as a measure of actual knowledge.
Defined the rubric and quality gates, built the agent pipeline that generates and scores questions, and shipped the offline practice app.
Calibrated the rubric against 37 exam-board sample questions and gated rebuilds on 12 numeric checks, catching 89.7% answer-position bias and systematic option-length tells.
Teams transcripts captured important decisions and actions, but left them buried in recordings and unusable by the workflows that followed.
Built the transcript capture and processing pipeline, with structured notes, routing rules, ambiguity flags, human review and downstream action handling.
Tested repeat processing and compared independent AI outputs, exposing duplicate transcripts and under-specified formatting rules. Fixes went into controls and specifications rather than blind model tuning.
Weekly competitor research was manual and inconsistent, producing large volumes of information without reliably identifying the strongest topic to pursue.
Built the eleven-channel research pipeline, scoring framework, data contract and validator that turn competitor activity into structured decision briefs.
Evaluated a delivered report against the original requirements and converted every identified gap into permanent rules or validators, with each validation rule tested independently.
Complex test-data queries often required five or six tables; unfamiliar joins could fail or create hard-to-explain duplicate results.
Built a Copilot-based natural-language SQL engine, onboarding table relationships, field meanings and business rules around policies, lives and benefits.
Used it on real test-data searches, checking generated queries against the intended conditions and distinguishing genuinely missing data from query-construction problems.
A 39-page website was being built rapidly with AI-generated code, where a small change could silently break unrelated pages or layouts.
Directed the AI-assisted build and defined what “correct” meant, turning design and quality requirements into automated checks and release gates.
A 52-check pre-publish gate runs alongside Playwright visual regression at four viewports, allowing intended changes while failing unexpected movement elsewhere.
A repetitive PAS anniversary workflow required hours of manual terminal navigation and data entry, with interruption creating duplicate, missed-work and recovery risk.
Built an Excel VBA automation using Reflection for screen observation and Windows key injection for actions, with controlled inputs, logging and rerun logic.
Validated compile, navigation, single-client, round-trip and batch behaviour in the live system; rerun testing correctly skipped 83 clients already recorded as successful.
Volkswagen was rebuilding its online car configurator within a wider digital programme spanning the customer configurator, dealer contact centre and CMS.
As Business Consultant, I gathered and maintained requirements, wrote user stories, worked with QA on review and testing, supported defect triage, prepared UAT journeys and reported to the Product Manager.
A new protection product required changes to TAL’s policy administration environment and close alignment between product rules, development and testing.
As Business Analyst for policy administration, I reviewed specifications, clarified requirements, wrote and maintained user stories, answered delivery questions and supported business UAT.
TAL introduced SailPoint to manage identity and access for a policy administration system used by multiple business teams with different permissions.
I mapped how teams used the system, translated those needs into 100+ access profiles, configured and tested them, and coordinated business UAT.
Claims functionality was migrated to Fineos while remaining integrated with TAL’s policy administration environment.
I focused on API, integration and system testing, including test data setup and verification that claims and status changes moved correctly across systems.
Connected audio products were tested across six countries using real combinations of devices, apps, ISPs, routers and user environments.
As Test Manager at Applause, I managed the distributed test programme, tester participation, execution, defect triage and client reporting across thousands of test cases.
I work best when I can bring clarity to complexity and help teams ship work that actually matters.
“I met Flo a year ago when we became their Odoo implementation partner. The horizon expanded quickly after that when we realised that our business thinking is alike. Since then, we have been collaborating with Flo on a couple of projects. He always asks the right questions that go deeper, comes up with out of the box ideas, and shares his learnings with us, which means we can go faster.”
“I didn’t think AI would make sense to use in an architecture practice, but we are now saving so much time with the automations he has established. It is hard to imagine where the world is going, but one thing is certain: it was the right call to talk to Florian to open our eyes as to the potential that is now at our fingertips! If you need slow processes optimised and sped up in your firm, make sure to give him a call.”
“As a builder, there are multiple things to be organised in parallel and sometimes the hardest part is making sure that my trades know what to do when, and getting the materials onsite when they need to be. Yes, there is plenty of software out there that can assist, but they all have a steep learning curve. Flo and the OIC team are building a tool that is going to make this much easier, and I am one of the first builders to try it out. I can definitely see how this tool will change the way I work!”
“Florian can look at a business process or a problem and come up with a technical solution. What interested us as a business was being able to have in-house capability to use AI to increase business efficiency. Florian taught us well, set up project initialisation procedures built around the way we work, and we can now use those learnings to move faster and provide our clients with a better client experience due to the improvements that were made.”
“When we talked to Flo, the first thing he asked us to do was run him through our business process. Flo was quick to identify opportunities for him to come in and provide value. What he ended up doing was build a financial feasibility tool for us that covers a range of client scenarios. This has allowed us to produce additional cashflow as an upsell opportunity! We are now looking at automating processes that have historically always been manual. After seeing what can be done, we are very confident that we will reach efficiencies that were previously not possible.”
“Flo built me a tool that automatically stores and organises every piece of information about a project. Everything that happens on the project is captured and organised automatically, so instead of relying on people to remember where information is stored, I can just ask the system what happened, what was decided or what I need to know. Everyone who runs a business and deals with dozens of clients every week needs this because we now spend less time looking for information.”
“When you run several social media platforms, the important thing is to be on top of what is happening in the industry, be it changes to regulations, major events, shifts in the market, etc, so that you can talk about it. I never really had the time to keep on top of things while running 2 businesses. Florian built me a social media research tool and intelligence matrix. He built it as per my vision, and now when I kick this off it gives me all I need to make a decision. This has allowed me to make content about what matters when it matters, while still being able to focus on delivering our services.”
“Florian is someone I have always trusted to get into the nitty gritty details and understand how a system or process really works. He is thorough and critical in the way he approaches his work and he always looks for ways to make things more efficient. I have enjoyed working with him immensely and, in fact, rehired him after he previously left TAL, which probably says more than anything about the level of trust I have in the quality of his work and the value he brings to a team.”
I help companies, from small firms to large enterprises, build software they can trust putting into production. That means understanding what the product is supposed to do, asking the right questions early, testing the things that actually matter, and making sure the people responsible for releasing it can say: we understand this system, we know how it behaves, and we are confident in what we are shipping.
I also help businesses use AI to do things that were previously too slow, too expensive or simply not possible. That can mean automating a process, building an AI-enabled workflow, or engineering a custom tool around the way the business already works. The goal is to give the team more capability without making them dependent on something they do not understand.
Yes. I work on contract engagements. I’m Sydney-based and open to onsite, hybrid and remote work. My availability changes, so send me the role, expected start date and working arrangement and I’ll tell you quickly whether I’m available and a fit.
Both. Building gives me direct exposure to the parts that are often messy at first: requirements, data, integrations, user behaviour and failure modes. My quality background means evaluation is considered while the system is being designed, rather than treated as something added at the end.
Start by defining the behaviours that matter, the unacceptable failure modes, where human oversight is required and what evidence would make the result good enough to use. Then evaluate those things repeatedly as prompts, models, data and workflows change.
I am most useful when things are messy — when there is not enough time, people have different ideas about what to do next, or a company knows it should be using AI but has no clear way to turn that into something useful.
I am also effective in teams that have realised quality needs more attention: an app is getting poor feedback, a product is becoming harder to maintain, features are not behaving consistently, or testing has become something that happens too late. I can help work out what is actually going wrong, ask the questions that uncover the real problem, and use a mix of experience, automation and AI to improve how the system is tested and delivered.
I’m tool-agnostic. In my own work I mainly use Python, JavaScript/TypeScript, Playwright, MCP, n8n, SQL, APIs and Git/GitHub. For a team, I would not force those tools into every problem. I look at the existing stack, risk, maintainability and governance needs, then choose or evaluate the right option. Depending on the context, that may include established testing platforms such as Tricentis or Katalon, AI-assisted tools such as Diffblue or Qodo, visual testing tools such as Applitools, or API tooling such as Postman. The tool is an implementation choice; the quality criteria come first.
Yes. Over the past two years I have helped dozens of colleagues improve the way they use AI in their day-to-day work; from getting better results out of the tools they already use, to setting up repeatable workflows, automating routine work and choosing the right AI approach for a particular problem.
I am very practical about it. I teach people how to get started, share the shortcuts and techniques that make a real difference, and help teams set up AI around the systems and processes they already have. The aim is not to make people “use AI more”. It is to give them useful tools they can actually work with themselves and benefit from every day.
