AI review

In two weeks, you’ll know whether your AI system holds up in production.

I check your initiative or your running system against five points and write down what you should go ahead with, adjust or stop.

2 weeks Fixed price Clear recommendation

The method

The Production Readiness Review.

Five checkpoints. For each one, the report gives a traffic light and a rationale.

This works just as well for initiatives that are still on paper.

01

Goal

Which problem should the system solve, and how will you know it is solved?

02

Architecture

Model, data connections and vendor dependencies. Plus the question of whether you should buy or build.

03

Quality (evals)

Are there test cases that show you whether a change makes the system better or worse?

04

Security

What happens when users or third-party content deliberately lead the system astray?

05

Operations

Cost per request, monitoring and ownership within the team once I am gone.

Experience

I only advise on what I can build and verify myself.

“Simon designed the architecture around our compliance agent KAIA and brought it to paying customers. He was instrumental in establishing evals and made the quality of our AI measurable. Simon takes on far-reaching responsibility, brings the team along and ensures the highest level of security and reliability, even on the most complex tasks.”

AP

Alexander Prams

CTO & Co-Founder, Kertos

Translated from German.

20+ years in software 5+ years full-time in AI

AllianzSiemensHeinrich-Böll-StiftungUniversität BernMANHochschule MünchenLufthansaHasso-Plattner-InstitutNokiaTactical TechEvonikStrascheg Center for Entrepreneurship (SCE)KertosVitolusAllianzSiemensHeinrich-Böll-StiftungUniversität BernMANHochschule MünchenLufthansaHasso-Plattner-InstitutNokiaTactical TechEvonikStrascheg Center for Entrepreneurship (SCE)KertosVitolus

My own company and the AI workflows of a research institute run on systems I built and operate myself.

What you get

A written report.
Plus a presentation.

Traffic light per checkpoint

An assessment with a rationale for each of the five checkpoints.

A clear recommendation

The report ends with an unambiguous recommendation: go ahead, adjust or stop.

Prioritised next steps

A list of what to do first and what can wait.

Results presentation

I present the results, to management or shareholders if you like.

How it works

Two weeks, three steps.

01

Understand Week 1

Conversations with three to five people and a review of systems, data and results so far.

02

Check Weeks 1 to 2

The five checkpoints of the Production Readiness Review, with code and test cases for running systems.

03

Recommend Week 2

Report and presentation with a clear recommendation: go ahead, adjust or stop.

What the review does not cover

No legal questions.

Whether a system complies with the GDPR or the EU AI Act is for lawyers to judge. I provide the technical facts (which data flows where, which vendors are involved) and a list of questions to ask your law firm.

No sales pitch.

The review ends with the report. You can carry on with it on your own, with your team or with another partner.

No 80-slide deck.

You get a document you can work with.

No courtesy diagnosis.

If AI is the wrong answer to your problem, the report says so.

And then?

If you want to keep building with me, this is the next step: Use-case lighthouse, 6 weeks to a running system →

Questions about the review

What does the review cost?

A fixed price, depending on scope. You usually get the concrete quote within five working days of the first call.

Do we need a running system already?

No. The review works before the investment just as well as for a pilot that does not make it into production.

How much of our team’s time does it take?

A few hours for conversations and access. I handle the rest independently.

Contact

Email me at info@slworks.de or book a free intro call right away.

Book an intro call

Loads the Cal.com booking calendar (USA). This transmits your IP address to Cal.com. More in the privacy policy (German).