A free introductory call — A paid engagement with a fixed scope.
5 working days of analysis, one written result, one fixed price. Afterwards you know what is slowing your delivery down and in which order it can be fixed.
The assessment is a paid engagement with a scope of its own — not a pre-sales call billed later, and not an open-ended discovery phase. We talk to your people, read your code and your pipelines, and write down what we find. On day five you get every result as files that belong to you. Whether you work with us afterwards is a separate decision.
If at least two of these sound familiar, the analysis is worth buying:
A release takes weeks, and only works when one particular person is available.
The cloud migration has been running for months and nobody can say when it ends.
The Kubernetes cluster runs, but nobody dares upgrade it.
The cloud bill grows faster than the product, and nobody can explain it line by line.
The person who understood the infrastructure has handed in their notice.
You are supposed to build a platform team and do not know what it should start with.
An investor, an acquirer or a major customer is about to run technical due diligence, and you would rather know first what they will find.
A commitment to the product or the market rests on a platform you are not sure will hold.
~/appetizer-labs/platform-assessment/fit
Who it is for
Engineering organizations with roughly 10 to 80 developers across two to eight teams.
CTOs, technical managing directors and platform leads who can approve a bounded engagement themselves.
Companies with their own development team — the results are written for the people who will act on them.
AWS, Azure, your own data centre, or hybrid. Kubernetes is often involved but is not a prerequisite.
Situations with a decision still open — rebuild, keep running, or replace — where nobody in-house is detached enough to make it.
Who it is not for
These enquiries get a no. Better you read it here than hear it after two meetings.
You are looking for 24/7 operations, on-call cover or SLA-backed run responsibility.
That does not exist here, and it will not exist after the assessment either. We would be promising something we cannot keep.
You need a certificate, an audit sign-off or a conformity statement — ISO 27001, NIS2, EU AI Act.
We are not a certification body or an accredited auditor. Engineering work towards a standard: yes. Attesting it: no.
You have no in-house development team.
Then there is nobody to hand the results to, and knowledge transfer turns into permanent outsourcing. That is a different business from ours.
You want a fixed-price quote for a whole rebuild without anyone having looked inside first.
Pricing an unexamined estate is guessing. That is exactly why this assessment exists — it is the step before, not a substitute for it.
Procurement runs as a formal public tender or only through existing framework agreements.
A founder-led company will not clear that process. The effort would be wasted on both sides.
You are looking for staffing, contractor placement or capacity at a day rate.
We sell an engagement with a result, not time on a rate card.
You want a second opinion that confirms a decision already taken.
We write down what we see. If that differs from the decision, it goes into the report the same way.
Your development team is well below ten people.
At that size the bottleneck is usually one you can name yourselves in a single conversation. That does not need a week of analysis, and it does not need a fee.
Six results, all written down, all in your hands on day five:
01
Bottleneck register
A sorted table rather than prose: every bottleneck found, with its effect, the order of magnitude of the effort, the risk of leaving it alone, and the evidence it came from — an interview, a repository, pipeline history or an incident report.
02
Architecture and delivery map
A diagram of the current state: systems, dependencies, and the path a change takes from commit to production — with the points marked where it waits instead of moving.
03
Risk and dependency map
Key people, expiring versions and support deadlines, single points of failure, and the dependencies between teams that never appear on an org chart.
04
Quick wins
Items your team can start without us and without a budget request, each with an order-of-magnitude effort and the effect to expect. Prioritized, so the list does not get parked as a whole.
05
Target state and 90 day roadmap
Where the platform should be in three months, in steps that can be delivered and reversed one at a time, with a named owner per step — names from your organization, not from ours.
06
Two readouts
A technical session with the team where every finding can be challenged, and a shorter version for the management or the board that states the consequence rather than the technology.
~/appetizer-labs/platform-assessment/timeline
The schedule, day by day
PreparationThe week before we start: access granted, documents shared, interviews scheduled. We read before we ask, so day one does not begin at zero.
Day 1Interviews, one person at a time rather than in a large round: development, operations, product, leadership. Separately, because a large round is where the necessary thing does not get said.
Day 2Architecture and code: repositories, infrastructure code, configuration. You walk us through a real change rather than describing one.
Day 3Delivery and operations: pipelines, environments, deployments, monitoring, and the most recent incidents. What interests us is where a change waits, not where it works.
Day 4Analysis without you: sorting findings, sizing effort, writing. By the evening you have a draft of the bottleneck register, so nobody is surprised in the readout.
Day 5Technical readout with the team, then the shorter version for leadership. You receive every file the same day.
The 5 working days usually sit inside 3 weeks rather than running back to back — your team needs to keep working in between, and we want time to read between the interviews and the code.
~/appetizer-labs/platform-assessment/inputs
What we need from you
Without these the analysis becomes guesswork. They are listed here so you know in advance what your side of it costs:
A named contact with decision authority who is reachable while the analysis runs.
Read access to repositories, infrastructure code and CI/CD configuration. Read-only is enough.
Read access or a shared screen for monitoring, logs and the cost view of your cloud account.
Six to ten people for one interview each, scheduled before we start.
Existing architecture and operations documentation exactly as it is. We do not expect it to be current — the distance to reality is a finding in itself.
The most recent incident reports or post-mortems, if any exist.
An NDA and, where personal data is involved, a data processing agreement. Both signed before the first access is granted.
You write a few sentences about what is stuck. You get an assessment back of whether this is the right instrument — and if it is not, what would be.
02
Scoping
One conversation to settle scope, dates and access. It is not a discovery phase and it is not billed; it establishes whether and when.
03
Order
Fixed scope, fixed dates, fixed price — in writing, before the first access is granted. No change requests for days that take longer than expected.
04
Delivery
The working days described above. Your side of the effort is bounded and known in advance: one interview per person, plus the readouts.
05
Handover
Every result as files you own: text as Markdown, diagrams as source, the register as a spreadsheet. Nothing sits in a portal you lose access to when the engagement ends.
06
Afterwards
Once the readout is done the engagement is complete. There is no obligation to continue and no follow-on contract in the small print. You can deliver the roadmap yourselves, deliver individual steps with us, or hand the results to another supplier — they are yours. If we think you need someone else for the next step, that goes in the report.
~/appetizer-labs/platform-assessment/pricing
What it costs
A fixed price for the whole scope, agreed in writing before we start. No day rate, no time-and-materials billing, no change requests.
Covered by the fixed price
Preparation, interviews and analysis
All six results as files
Both readouts, remote
Follow-up questions on the report for two weeks afterwards
Travel for one on-site day in North Rhine-Westphalia
The figure is not on this page yet. Send us your team size, your platform and where you stand — you get the fixed price back in writing, before a meeting is needed.
~/appetizer-labs/platform-assessment/example
What the result looks like
The bottleneck register is the core of the report. This is how it is built — one row per bottleneck, sorted by effect over effort:
Bottleneck
Effect
Effort
Risk of leaving it
Evidence
Release approval depends on one person
Releases wait for a calendar slot rather than for finished code
M
High — if that person is unavailable, all delivery stops
Interviews 3, 4 and 7; pipeline history
The row above is an illustrative pattern, not a client case. We do not publish results from real assessments, anonymized or otherwise.
Effort is given in orders of magnitude (S, M, L) rather than person-days: person-days estimated out of a single week of analysis would be guesses, and a guessed number in a table looks exactly like a measured one. The roadmap below it arranges the same rows into steps for the next 90 days.
~/appetizer-labs/platform-assessment/next
What can follow
The roadmap names the steps, not the supplier. If you want to deliver some of them with us, these are the services an assessment most often leads into — each bookable on its own, with nothing bundled:
No. The assessment is an engagement of its own with a fixed scope and a fixed price, and it is complete once the readout is done. There is no obligation to order anything afterwards. The results are yours, including if you hand them to another supplier to deliver.
Why 5 days and not two weeks?
Because 5 working days are enough to find the bottlenecks, and two weeks would start solving them. Solving is delivery, is ordered separately, and should not run quietly inside an analysis.
Who actually does it?
Matthias Bruns, founder of Appetizer Labs. There is no team it gets passed to internally: the same person runs the interviews, reads the code and writes the report.
What if you find nothing?
Then that is what the report says, with the reasoning, and you have a defensible answer instead of a suspicion. We do not add items to make the list longer.
How much of our time does it take?
About an hour per interviewee, plus the two readouts. The named contact also spends time preparing access and scheduling — in total well under a working day, spread across the three weeks.
Remote or on site?
Remote-first; interviews and readouts work well remotely. One on-site day in North Rhine-Westphalia is included in the fixed price; outside NRW, travel is charged separately.
In German or in English?
Either. You choose the language for interviews, report and readouts — including a mix, such as an English report and a German readout.
What happens to our access and documents afterwards?
Read access is revoked after the readout, and shared documents are deleted or returned. The NDA and, where needed, the data processing agreement come first.
05
~/appetizer-labs/contact
Request the assessment
Send us your team size, your platform and what is currently stuck. You get the fixed price and available dates back in writing — and an honest no if an assessment is the wrong instrument for you.