A factual diagnosis of your product and your tech.
Deliveries that slip, bugs piling up, AI analyses no one knows what to do with. In three weeks, an assessment that cross-references your data with what your teams experience, so you stop making trade-offs on instinct and know where to act.
From blur to factsEveryone has their perception. No one has the facts.
It's the third quarter in a row that the same feature slips
The start-of-quarter roadmap never holds. At every steering committee, we adjust, postpone, re-explain. Teams end up no longer believing in it, and neither does the business side.
We have dashboards and AI analyses, but we get nothing out of them
Metrics, AI-generated reports, there is no shortage of data. But no one knows what matters, what is noise, or what it really says about your teams.
Several people think they decide, no one owns it
Trade-offs get made three times, in three places, in three different versions. Decisions do not hold.
We run tech like a cost centre, and expect unicorn results
The team is under pressure on deadlines and budget. You sense the contradiction, without knowing where to cut through it.
From impressions to shared facts.
From perceptions to facts
The teams' feelings are set against your real data (tickets, lead times, deliveries), analysed in depth by my AI tools.
The structural from the cyclical
Deep-rooted causes (legacy, processes, organisation) are separated from passing difficulties. You know where to invest.
A shared basis for discussion
A diagnosis shared across leadership, product and tech. Decisions rest on facts, no longer on impressions.
Two formats, depending on your need.
The audit comes in two scopes. What comes after (an engagement, training, a hire) is defined once the diagnosis is set, never before.
Product & Tech audit
The organisation, the product, the tech and quality examined together. An approach built on Marty Cagan's product framework and DORA metrics.
- 10 to 15 interviews, one to one (45 min)
- Anonymous questionnaire (~60 questions)
- Observation of your rituals
- Report: diagnosis + recommendations
- 2-hour debrief to leadership
Who it's for: several issues stack up (product, organisation, quality) and you want an overall read before acting.
Audit of a specific pain point
A single topic explored in depth: bug management, velocity, roadmap prioritisation, quality, the product-tech relationship. Diagnosis and action plan on that scope.
- Interviews focused on the scope
- Analysis of the relevant data
- Diagnosis + action plan
- Debrief of the findings
The number of interviews and the duration are set to the breadth of the topic, during scoping.
Who it's for: you have identified the symptom, you are missing the diagnosis and the plan.
The audit takes your regulatory constraints into account (IEC 62304, MDR, ISO 13485). The recommendations stay compatible with your quality system, without freezing your teams.
Cross-referencing what the teams say and what the data shows.
During the audit, your teams keep working as normal.
One-to-one interviews
45 min per person, from leadership to developers. Anonymised verbatims.
Observation of rituals
Sprint planning, daily, demo, committee. The organisation as it really works.
Anonymous questionnaire
Around sixty questions to quantify what the interviews reveal.
AI analysis of your data
Tickets, lead times, flows, documentation, combed through by my AI analysis tools.
Putting it in perspective
What the data shows, set against what your teams say, and put in perspective within your context: your market, your constraints, your history. It is this cross-reading that makes the diagnosis, not the numbers alone.
This is what the debrief looks like.
This one from a focused audit on bug management. Client anonymised, figures changed, but the structure and the approach are those of the document delivered.
- Executive summary · readable in five minutes
- Quantified findings · each chart with a clear reading
- Strengths to build on · what already works
- Actionable recommendations · finding, actions, benefits
An audit is only worth the decisions it makes possible.
Two recent examples.
A decision to end the part-time CPTO engagement and hire a CTO in-house. The audit served to frame the profile and prepare the handover.
With the diagnosis and priorities set, the teams did the rest on their own: bug backlog cut in half in six months, fixes finally aimed at the old bugs. Results measured on the data, by re-running the audit six months later.
“Thank you for this in-depth analysis, highly relevant and detailed. I can see the work it represents. I broadly agree with your conclusions, and several suggestions are worth implementing.”
Montaine Marteau
I run every audit myself, end to end, not a junior consultant. Engineer and PhD by training, more than twenty years in Tech and MedTech, most of it in product leadership. From the field to the exec committee, I read an organisation at every level: product, tech, business, leadership.
I am not a developer: I audit the organisation, the product and the delivery. And increasingly your AI: in each person’s productivity, in your team processes, or in the product itself.
Where do you really stand?
F.A.Q.
Answers to the questions I get asked most. Another one in mind? Ask it during the call.
Will it disrupt my teams?
No. It is 45 minutes of interview per person and the observation of rituals that are already scheduled. Your teams keep delivering during the audit.
Are the interviews confidential?
Yes. All verbatims are anonymised. No one, leadership included, can link a comment to a person.
Am I obliged to follow on with an engagement?
Never. The audit is a standalone deliverable, actionable by your teams alone. If what follows goes through me, it is priced separately, after the diagnosis.
How much of my own time does it take?
A one-hour scoping session at the start, your interview, the debrief. The rest runs without calling on you.
What access do you need?
At a minimum, read-only access to your tracking tools (tickets, delivery) and to your rituals. The more access you open (code, documentation, product data), the deeper the analysis can go. In every case, it is read-only, never modification.
Could we do this analysis ourselves, with AI?
The raw analysis, yes, and I encourage you to. What you buy here is not computing power. It is knowing what to ask the AI, the cross-referencing with the interviews and the rituals, and an outside reading that no one inside can carry, because inside everyone has a stake in the outcome.
And if the audit is not the right answer?
We see that during the 30-minute call. If an audit is not relevant for you, I tell you so and point you in the right direction.
Unsure about your product or tech organisation?
Book 30 minutes. You lay out your situation, we look together at whether an audit is the right answer, and which one.