If you are considering AI in an internal process or your own application, I help connect the business brief with user needs, data and technical options. I may also recommend a simpler solution or holding off on AI.
AI strategy and project management
I help you assess whether your AI idea is suitable for day-to-day operations. I prepare a concept and the information you need to decide whether and how to proceed. If we agree to implementation, I can manage the project with your team and suppliers. We define the scope around your brief.
Discuss your AI projectDecide first, then implement
A one-off consultation helps with a specific decision. Project leadership is a separate follow-on engagement: from requirements and priorities through supplier selection to reviewing outputs and handing over to your team. It is not an automatic commitment to the entire development project or an offer of a standalone development team.
AI: test the use case before investing further
Problem, data and constraints define the brief
How we decide on the next step
Sometimes a simple automation or process change is a better fit. If we choose AI, we test it within a limited scope first. We agree on further work based on the findings.
- Process and peopleProblem, users, data and an owner.
- A bounded testOptions, risks and acceptance criteria.
- Agreed deliveryPriorities, suppliers and handover.
Evidence and idea assessment
What we need for the assessment
- The current process, problem and person accountable for the outcome
- Users, workload and the consequences of errors
- Available data, its quality, accessibility and access restrictions
- Budget, timing dependencies and internal team capacity
- Previous experiments, supplier proposals and expectations of AI
Where AI fits and what to test
We compare possible uses of AI by expected value, feasibility, risk and the effort needed to change the process. We distinguish what the data supports from what a pilot still needs to test. I do not promise time or cost savings in advance: first we need a baseline and agreed measurement conditions.
Solution concept, technology and roadmap
Process and business objective
Technology options
Option assessment → pilot proposal and roadmap
What I prepare for the decision
- A solution concept tied to a specific process and business objective
- Assessment of technical feasibility and open questions
- Comparison of technology options: an existing tool, custom solution or a combination
- Data, integration, permission and human-review requirements
- A pilot plan, acceptance criteria and quality, cost and usage metrics
- A roadmap with priorities, dependencies and decision points
How the work can continue
Project leadership and supplier oversight
A follow-on engagement
If you decide to continue, we can move on to preparing a brief, comparing proposals and coordinating delivery.
Coordination and oversight
I review whether outputs meet the brief and keep priorities, responsibilities and dependencies clear between management, your internal team and suppliers.
Agreed scope
Within the agreed scope, I work on the backlog, output reviews, user feedback, documentation and operational handover.
Not automatically included
Development, operational support and specialist involvement need separate agreement. We establish who provides them and where their responsibilities begin and end.
Agents, automation and company knowledge
Options around the problem
Depending on the problem, we can assess a simple automation, an assistant over company data, a RAG system or a multi-step agent.
Shared conditions
These are solution options, not a mandatory technology package. For each, we define permission boundaries, human review, answer validation and failure handling.
Internal use, not brand visibility
This service concerns internal AI use. Brand visibility in search engine and chatbot answers is a different task, addressed through GEO and SEO.
Data, security and responsibilities
Data and operations
The brief defines how to handle sensitive data, access, audit trails, monitoring and failures.
Accountable roles and legal limits
I review GDPR, security and operational requirements with the people responsible for those areas in your organisation. This is not a legal opinion or a guarantee of regulatory compliance.
Suppliers and portability
We also compare supplier dependency, data portability and, where relevant, private or local hosting.
Limit of independence
Complete independence from suppliers cannot be promised without knowing the chosen solution.
How to start
Send a short description of your process, the problem and what you have already tried. We will discuss whether a one-off assessment, strategy preparation or ongoing project leadership makes sense. We confirm specific outputs, technical specialist involvement, price and responsibilities before work starts; a developer does not need to attend the first consultation.
Engagement scope
Assessment and implementation are separate decisions. We move to the next phase only on the basis of findings and agreement.
- Assessment:
- Process, feasibility, solution concept, technology options and proposed next steps according to the brief.
- Project leadership:
- Priorities, requirements, supplier selection and coordination of outputs within the agreed scope.
- Prerequisites:
- Available information, an accountable owner and cooperation from your internal team; technical validation may require a specialist.
No advance promise of savings, return on investment or a successful pilot. This service does not replace a development team or legal and security approval.
Context and next steps
Go to consultation topics for AI projects.
Options for automating a specific process and the limits of AI involvement.
How I connect visibility in AI answers with ongoing SEO.
Frequently asked questions
Do we need to choose a technology first?
No. We start with the process, data, expected value and risks. Only then do we compare suitable technologies and ways to test them.
Is a one-off consultation enough?
It may be enough for a specific decision. Strategy and implementation leadership need a broader brief and a separately agreed scope. A consultation does not commit you to continuing the project.
Do you also deliver custom applications?
I can help with requirements, priorities, supplier selection and validating outputs. Specific development work and its owners are agreed per project; this is not a promise to develop the entire application myself.
How do we assess whether a pilot works?
We define the baseline, test scenarios and acceptance criteria beforehand. Depending on the process, we track output quality, time, cost, error rates and user adoption. A successful pilot is not guaranteed.
Does a developer need to attend the first consultation?
No. To start, we need someone who knows the process and can explain the brief. A technical specialist can join depending on the questions that need validation.
From an idea to a testable prototype.
I can help clarify an idea, prepare a prototype or lead development of a complex AI application. Depending on the project, I work independently or with colleagues specialising in AI development; this is not a promise of an in-house development team. We clarify responsibilities, required data, permissions and what can be sent to external tools beforehand. I consider GDPR, commercially sensitive information and anonymisation where appropriate.
Have a specific AI project in mind?
Describe your process and the decision you need help with.