AI consulting for Swiss companies

Whether you do not yet know where to start, have twenty ideas without an order, or a pilot that will not scale — we start where you actually are.

Our understanding of consulting

Consulting with an obligation to enable

AI turns operational work into a commodity. What remains is judgement — and the question of whether a consultancy wants to tie you in operationally or make you more independent.

We build every engagement to leave the knowledge and the structures inside your company that you need to continue without us. That is not a website posture: it is in the outcomes of every single service — and, since 6 August 2026, an obligation to enable in clause 2 of our terms of business.

We also only advise on things we have built ourselves. What we say about RAG systems, agents or AI compliance comes from running them in production — not from a market study.

Where do you stand right now?

“We have nothing yet, and nobody knows where to start”

We develop a roadmap with your leadership team: which use cases first, which later, where to invest. The result is a justified sequence with quarterly milestones — and a watchlist of what is consciously deferred.

AI Strategy

“We have twenty ideas and no order”

A discovery process with your business units and a prioritisation matrix of effort, value, strategic relevance and risk. The result is a piloting recommendation you can defend internally.

Use Case Identification

“We have to choose a vendor and do not know by what criteria”

We write the requirements document with you before you talk to vendors — not afterwards. Plus market analysis, RFP, neutral evaluation and contract negotiation, with particular attention to Swiss nDSG, GDPR and EU hosting.

Vendor Selection & Requirements

“We bought something, and it is not going live”

That is rarely a product problem. We treat AI adoption as a change process: stakeholder management, internal communication, training, and success measurement that shows whether usage actually emerges.

Change Management & Training

“Our pilot runs, but it does not scale”

The most common point at which AI projects stall — and it usually comes down to architecture and operational maturity, not the model. We work along technically: architecture reviews, code reviews, sparring with your engineering teams, quality gates.

Implementation Support

“Our customer service is drowning in standard enquiries”

We draw the line between enquiries AI can answer and those that belong with people — and build the handover in between. We work on your anonymised ticket data, not on industry examples.

Conversational AI & Customer Dialogue

“Compliance is breathing down our neck”

Swiss nDSG, EU AI Act 2024/1689, GDPR. Gap analysis, policy development, building a compliance programme and audit preparation — particularly relevant for finance, pharma and the public sector.

AI Compliance & Governance

How we differ

Most consultancies in this market are either large houses with a methodology deck, or freelancers with one specialism. We are neither, and the difference matters for what you should expect from us.

  • We build what we advise on. Nexbid, Mingly, DocMind, RAG-Wissen and Prüfstand are our own production systems. When we say a RAG architecture breaks at a certain point, that is from operating one — not from a vendor briefing.
  • We are small, and we say what that costs you. You get the people who did the work, not a pyramid. In exchange we cannot staff a twenty-person programme, and we will say so rather than stretch.
  • Our engagements end. Every service lists the work products that stay with you, and clause 2 of our terms of business binds us to hand over the knowledge to continue without us — where a large house would see a renewal.
  • We are Swiss-based and Swiss-regulated. nDSG, EU AI Act and EU hosting are not an add-on chapter for us — they decide which architectures we recommend in the first place.

Where you can check this

The consulting services in detail

AI Strategy

We develop a clear AI roadmap with your leadership team: which use cases first, which later, and where to invest.

Includes

  • · Roadmap workshops
  • · Prioritization matrix
  • · Investment recommendation
  • · Quarterly review

How it runs

  1. Leadership kickoff workshop: where the company stands, what the pains are, what the hypotheses are
  2. Discovery in business units: collect use cases (typically 15-30 candidates)
  3. Prioritisation matrix along business value × data maturity × feasibility × change effort

You take away

  • Shared leadership vocabulary and prioritisation logic
  • Three use cases with investment indication that you can run without us
  • Clear watchlist for later (what consciously not now)
  • Quarterly review cadence you orchestrate yourselves
Learn more and book

Use-Case Identification

Discovery sessions in business units uncover the use cases that actually create value — not the ones that look good in slide decks.

Includes

  • · Department workshops
  • · Use-case backlog
  • · ROI estimate
  • · Pilot proposal

How it runs

  1. Pre-discovery: review existing data landscape, understand IT architecture
  2. Discovery workshops with each business unit (typically 2-4 half-days)
  3. Pain-point mapping: where employees spend their time, where repetition lives, where frustration

You take away

  • Prioritised use-case backlog for 12-18 months
  • ROI estimate per use case with traceable calculation
  • Concrete pilot proposal — no theory
  • Internal vocabulary for evaluating further use cases
Learn more and book

Vendor Selection & Requirements

We write the requirements document, evaluate vendors neutrally, and support negotiations — so you don't end up buying what sells best.

Includes

  • · RFP creation
  • · Vendor scoring
  • · Contract negotiation
  • · Reference calls

How it runs

  1. Requirements workshop: functional, technical, regulatory, and commercial
  2. RFP creation with traceable weighting of criteria
  3. Long-list vendor research, short-list to 3-5 vendors

You take away

  • Traceably documented selection process (audit-ready)
  • Contract with fair SLAs and clean exit clause
  • Realistic effort and cost expectations
  • Internal knowledge for future own vendor evaluations
Learn more and book

Implementation Oversight

While external developers build, we ensure quality gates, technical oversight, and a clean handover to your team.

Includes

  • · Architecture reviews
  • · Sprint accompaniment
  • · Quality gates
  • · Handover program

How it runs

  1. Architecture review at project start: does the stack make sense, are risks addressed
  2. Sprint accompaniment as the client's technical sparring partner (not the supplier's)
  3. Quality gates per sprint: code quality, test coverage, privacy, performance

You take away

  • Technically clean implementation without vendor dependency
  • Internal champions who evolve the system without us
  • Audit-ready documentation trail
  • Reduced need for external support over time
Learn more and book

Change Management & Training

Most AI projects fail not on technology but on adoption. We develop internal champions and run follow-up training sessions.

Includes

  • · Adoption programs
  • · Champion selection
  • · Follow-up trainings
  • · Feedback loops

How it runs

  1. Stakeholder analysis: who are the advocates, who the skeptics, what are the concerns
  2. Champion program: selection and development of internal multipliers
  3. Training series in waves — not all at once, but in iterative cohorts

You take away

  • Adoption as a measurable quantity, not a gut feeling
  • Internal champion structure that carries follow-up use cases
  • Training material that stays with the company
  • Realistic adoption expectation and escalation paths for problems
Learn more and book

Conversational AI & Customer Dialogue

We draw the line between enquiries AI can answer and those that belong with people — and build the handover in between.

Includes

  • · Enquiry classification by complexity and AI suitability
  • · Tooling concept: classification, draft responses, FAQ bot
  • · Vendor selection and Swiss-nDSG data protection impact assessment
  • · Training on prompt adjustment and escalation paths

How it runs

  1. Enquiry inventory: we sort the real ticket volume by frequency, effort and escalation risk
  2. Drawing the line: which enquiry types suit automation, which must stay with people
  3. Tooling concept: automatic classification, LLM draft responses with human approval, FAQ bot for standard questions

You take away

  • Sorted enquiry inventory with an automation recommendation per type
  • Tooling concept with a clear human/machine boundary
  • nDSG assessment for customer data in prompts
  • Handover concept and escalation paths the team maintains without us
Learn more and book

AI Compliance & Governance

nDSG, EU AI Act, and internal policies for responsible AI use — pragmatic and audit-ready.

Includes

  • · Privacy impact assessment
  • · AI Act classification
  • · Internal policies
  • · Audit preparation

How it runs

  1. Inventory of all AI applications in the company (often more than expected)
  2. AI Act classification: which risk class each application belongs to
  3. Privacy impact assessment for high-risk applications under Swiss nDSG

You take away

  • Complete overview of the AI landscape in the company
  • Compliance status per Swiss nDSG and EU AI Act
  • Audit-ready documentation
  • Internal policies employees actually use
Learn more and book

AI Crawler Audit

We audit robots.txt and Cloudflare settings across all your domains, document the state before the deadline, and record a reasoned decision per domain — instead of one blanket rule for all.

Includes

  • · Audit of robots.txt and Cloudflare settings per domain
  • · Baseline documentation before 15 September 2026
  • · Bot classification by effect, checked against operator documentation
  • · Reasoned decision per domain, in writing
  • · Follow-up check in Search Console and server logs

How it runs

  1. Inventory: every domain behind Cloudflare, plus its current robots.txt verbatim
  2. Secure the baseline: screenshot of bot settings, crawl statistics export, robots.txt archived — each with a date
  3. Classification: which bot sits in which effect class, and what a disallow there actually costs

You take away

  • For every domain it is documented what it permits AI providers to do — and why
  • The state before the deadline is provable, not remembered
  • No training block that secures nothing because the same content sits open elsewhere
  • Not for you: if you are not behind Cloudflare and allow model training anyway, there is nothing to do and you do not need this audit
Learn more and book

And if AI systems cannot find you at all?

A separate topic that fits none of the situations above: whether ChatGPT, Claude, Perplexity and Google are even allowed to read your content — and what your robots.txt actually permits today.

See the AI crawler audit

New · Advice by subscription

The AI Office Hours

from CHF 900 / month

Unlimited written requests, an answer within one or two working days, a monthly call. No quote, no budget request, no hour counter.

See packages and prices

Frequently asked questions about AI consulting

What does a consulting engagement cost?
We deliberately state no flat rate. Consulting is billed by effort, and the effort is stated per service in consulting days — for example 6 to 10 days for use case identification, 8 to 15 for vendor selection, one to two days for the AI crawler audit. In the discovery call we estimate your specific case before anything is signed.
How long does an engagement take?
Between one to two days and fifteen months, depending on the service. An AI crawler audit is done in one to two days, use case identification is complete after four to eight weeks, implementation support typically runs across the entire implementation phase of four to twelve months, and an AI strategy engagement runs three months of development plus twelve months of quarterly accompaniment. The figure is stated per service.
Do we have to book a workshop first?
No. Many clients start with a workshop because the risk is contained — but it is not required. If a workshop leads to a consulting engagement of CHF 8,000 or more within six months, we credit half the workshop price.
Do you work with small companies too?
Yes, that is our core audience. For smaller undertakings there is the AI Office Hours monthly subscription from CHF 900 — ongoing assessment and recommendation without a full engagement. Below that, a workshop is usually the more suitable format.
Do you also implement, or do you only advise?
Both, but billed separately. Consulting is guidance and judgement; development is its own service line with its own project pricing. In the discovery call we tell you which of the two actually solves your problem — even when it is the smaller one.
What happens after the engagement?
Ideally nothing you do not order. Every service ends with work products that stay in the company — roadmap, requirements document, policy, measurement setup. If you want continued support, there is the AI Office Hours subscription or a three-month enablement engagement in which delivery stays with your team.
Do you also advise in English?
Yes. Engagements run in German or English, and we deliver documents in the language in which decisions are made at your company. With international leadership teams that is usually English, even when operations are German-speaking.

Book a discovery call

30 minutes free. We discuss your context and propose a suitable scope.

Proof of practice

We advise on topics we have built ourselves.

See our own systems →