Service · typically 4–6 weeks
AI strategy that turns into shipped systems.
Most AI strategies are slide decks. Ours is a ranked list of workflows worth automating, a target architecture that fits your systems, the guardrails you need under GDPR and the EU AI Act, and a business case with numbers you can defend. Then we help you build it.
Why a strategy first
The expensive mistake is not choosing the wrong model. It is automating the wrong process, in the wrong order, without a way to measure it.
Language models and AI agents can now read, write, decide and act inside your systems. That makes almost every process a candidate, which is exactly the problem: teams start with whatever demo looked good, pilots multiply, and nothing reaches production because nobody agreed on value, ownership or risk.
An AI strategy fixes the order of operations. We map where time and money actually go in your company, score the candidate use cases on value, feasibility and risk, and sequence them into a roadmap that starts with quick, visible wins and builds toward the workflows that change your cost structure.
The result is deliberately short. Executives can read it in an hour, and engineers can start building from it the following week.
- FormatWorkshops, interviews, analysis, written roadmap
- DurationTypically 4–6 weeks
- ForFounders, boards, C-level, department heads
- OutcomeRanked use cases, architecture, governance, business case
What you get
Every engagement ends with the same five artefacts. They are written to be used, not filed.
Use-case portfolio
Twenty to forty candidate workflows across marketing, sales, operations, finance, HR and more, each scored on value, effort, data readiness and regulatory risk.
Prioritised roadmap
Three horizons: what to ship in the next 90 days, what to build over 12 months, and what needs data or organisational groundwork first. With owners, dependencies and decision points.
Target architecture
Which models, agent frameworks and platforms fit your stack, where data lives, how agents connect to your ERP, CRM and ticketing systems, and what stays in the EU.
Governance model
Roles, approval paths, human-in-the-loop checkpoints and a risk classification under the EU AI Act, so legal and works council questions are answered before they block a launch.
Business case
Hours saved, cycle times, error rates and revenue effects per use case, with the assumptions written down, plus the running costs of models, platforms and operations.
Operating model
Who owns AI in your company, how teams request and evaluate new agents, what skills you hire for and what you buy in. Small companies get a one-page version.
How it works
Discovery
Interviews with leadership and the people who actually run the processes. We read your process documentation, look at your systems and data, and collect the pain points nobody has written down.
Opportunity mapping
Two working sessions per function to translate pain points into candidate use cases. We show live examples of agents doing comparable work so the discussion stays concrete.
Scoring and sequencing
Each use case is scored on value, feasibility, data readiness and risk. We validate the top candidates with rough prototypes where the feasibility is uncertain.
Architecture and governance
We design the target architecture and the governance model against your existing IT landscape, security requirements and data-protection obligations.
Board-ready roadmap
A written roadmap and a leadership session where we walk through the decisions, the trade-offs and what happens in the first 90 days.
Who this is for
Founders who need to decide where AI belongs in the product and the company
Startups rarely lack ideas. They lack a defensible sequence. We help you separate what makes the product better from what makes the company cheaper to run, and fund both sensibly. See how we work with startups.
Mittelstand leadership under pressure to "do something with AI"
Owner-managed companies want a plan that fits their people, their ERP and their margins, not a transformation programme. We deliver a roadmap that a lean IT team and one committed department head can execute. See Mittelstand.
Corporate functions that already have pilots but no production
If you have a dozen proofs of concept and a steering committee, the missing piece is usually governance, ownership and a reference architecture. We provide all three. See enterprise.
What makes our strategies different
Written by people who build
The same consultants who design your roadmap can deliver the agent pilots and the workflow automation that follow. Nothing in the roadmap is there because it sounded good in a workshop.
Vendor-neutral
We have no reseller agreements. Whether you end up on Microsoft, Google, AWS, OpenAI, Anthropic, Mistral or open-weight models on EU infrastructure depends on your data, your systems and your risk appetite.
Governance included, not appended
Data protection, the EU AI Act, works council involvement and information security are handled inside the roadmap. Read more about our AI governance service.
Measurable by design
Every use case has a baseline and a target metric before anyone writes a prompt. That is what later lets you say whether it worked.
Frequently asked questions
How is an AI strategy different from an AI readiness assessment?
The readiness assessment is a shorter diagnostic: where do you stand on data, systems, skills and governance, and what are the two or three obvious first use cases. The strategy goes further: a full use-case portfolio, a target architecture, a governance model and a business case, sequenced into a roadmap. Many clients start with the assessment and extend into a strategy.
Do we need a lot of data or a data science team first?
No. Most agentic use cases run on the documents, emails, tickets and ERP records you already have. Where data quality genuinely blocks a use case, the roadmap says so and includes the data foundations work needed. You do not need a data science team to start; you need one process owner who cares.
Which departments should we start with?
The ones with high volumes of repetitive, document-heavy or communication-heavy work and a measurable outcome: customer service, order processing, accounts payable, sales development, HR helpdesks, contract review. The strategy ranks your specific candidates rather than applying a generic list. Our solutions by department pages show typical starting points.
How do you handle GDPR and the EU AI Act in the strategy?
Each use case gets a preliminary risk classification and a note on the personal data involved. The governance model defines who approves what, where humans stay in the loop and how you document decisions. This is not legal advice, but it means your legal and data-protection teams get a structured input instead of a surprise.
Can you also implement the roadmap?
Yes. That is the point. Our agentic workflow and agent development services deliver the first use cases, and managed AI operations keeps them reliable afterwards. You are equally free to implement with your own team or another partner; the roadmap is written so that you can.
How much time does our team need to invest?
Plan for roughly two to three hours per week for the sponsor and about the same for each department head involved during the mapping phase. We do the analysis, the writing and the prototyping.
Related services
AI readiness assessment
A two- to three-week diagnostic across data, systems, skills, governance, processes and culture, ending in three first use cases and a 90-day plan.
Agentic workflow automation
Multi-step business processes automated end to end by LLM-powered agents across ERP, CRM, ticketing and email, with human approval steps built in.
AI governance, EU AI Act & GDPR
An AI register, risk classification under the EU AI Act, GDPR-aligned processes and a usage policy your teams will actually follow, built together with your lawyers and your data protection officer.
Fractional Chief AI Officer
A senior AI leader embedded in your company one to two days a week: owns the roadmap, decides on vendors and architecture, runs governance and reports to the board, until you no longer need one.
Let's find the first workflow worth automating.
A 30-minute intro call, no slides and no obligation. We listen, ask about your processes, and tell you honestly where AI agents would pay off and where they would not.