AI & automation

AI that works with you. Processes that run by themselves.

Data is entered twice, enquiries sit waiting, knowledge is locked in heads and folders. We connect your systems, automate clearly defined processes and use AI where text, documents or knowledge need processing: with checks wherever a person should decide, and with data in Switzerland where required.

  • Automation
  • AI assistants
  • Documents
  • Knowledge

Challenges

Challenges
we solve.

Typical situations from daily work and how we approach them.

01

The same data is typed into three systems.

How we solve it

An automated process takes the data from a form, an email or the CRM and creates it in the other systems. Errors and duplicates are detected and reported instead of passing silently.

More on this in our blog
02

Email enquiries sit waiting because someone has to read, classify and look things up first.

How we solve it

An AI assistant reads new emails, recognises the request, gathers the information it needs and prepares a draft reply. Nothing is sent until your team has checked it.

More on this in our blog
03

You want to use AI, but client data does not belong in public chat tools.

How we solve it

We run AI in a protected environment, with data in Switzerland where available. Your inputs are not used to train the models, and access runs through your company accounts.

More on this in our blog
04

People don’t trust the AI, and rightly so, as long as nobody checks its output.

How we solve it

We test every solution before it goes live with real examples from your work, measure how often it gets things right and agree what may run automatically and what needs approval. Every step is logged.

Offerings

Our offerings
in detail.

Each offering can be booked on its own or combined with others. We agree the exact scope together.

Automate processes4

Automate business processes

One process across several systems, traceable and with clear exceptions.

  • Connections to the applications involved
  • Approvals wherever a person should decide
  • Error handling, logging and documentation

Approvals and requests

Holidays, expenses or purchases: requests run digitally, and their status is visible at any time.

  • A form instead of email ping-pong
  • Approval with one click, on the move too
  • A traceable history of every request

Connect systems

CRM, accounting, web shop and line-of-business applications exchange data instead of someone retyping it.

  • Clarification of the interfaces and data fields
  • Synchronisation in both directions where needed
  • Checks for duplicates and errors

Simple business applications

Excel lists and paper forms become applications for the office and on the move.

  • A data model and interface built around your process
  • Roles, permissions and validation rules
  • Your existing data brought over

AI at work5

AI assistants for your team

AI right inside the programs your team already uses: drafting text, summarising meetings, finding information.

  • Preparation of permissions and data
  • A pilot with tasks from your daily work
  • Training and rules for working with AI

Email and enquiry assistant

Classify incoming enquiries, gather information and prepare replies.

  • Classification by request, urgency and owner
  • Draft replies based on your templates and data
  • Sent only after approval

Document processing

Capture information from receipts, orders or contracts reliably and pass it on.

  • One defined document type with the fields you need
  • Validation rules and review of uncertain cases
  • Hand-off to the destination system

Knowledge assistant

Answers from your manuals, templates and policies, with sources to check.

  • A defined knowledge collection and access rules
  • Answers with references to their sources
  • Evaluation with questions chosen together

Assistant for client enquiries

An assistant on your website answers frequent questions from approved content and passes the rest to your team.

  • Answers only from content you approve
  • Handover to a person, with the conversation so far
  • A privacy notice and a log of conversations

Foundations3

AI policy and training

Clear rules for using AI in the company, and a team that knows them.

  • A usage policy: permitted tools, data and review duties
  • Support with the data protection assessment, without legal advice
  • Training on using AI safely

A secure AI environment

Your own AI solutions in your environment, with data in Switzerland where available, protected access and monitored costs.

  • The right model, chosen by quality, confidentiality and cost
  • Access through your company accounts only
  • Logging, budgets and monitoring

Data and reporting

Key figures from different systems in a report that updates itself.

  • Clear definitions of the figures
  • Dashboards and regular reports
  • Traceable data sources

An example project

From enquiry to order.

An enquiry arrives by web form or email. The contact is recorded in the CRM automatically, AI summarises the request and a quote is prepared as a draft. Once you approve it, it goes out, with a reminder if no answer arrives.

  1. 01Enquiry
  2. 02Draft
  3. 03Approval

Frequently asked questions

Good to know.

Your question is not here? Write to us

Which tools do you use?

That depends on your environment. If you work with Microsoft 365, we often use the automation and AI already included in it. For many different systems or special requirements, we use other proven tools. What matters is that your team can use the solution and carry it forward.

More on this in our blog
Is our data used to train the AI?

We choose AI services so that your inputs are not used to train models, and we check the terms for each use case.

How reliable is AI?

AI makes mistakes. That is why we test with real examples, check results against rules and let a person decide anything uncertain. Only what has proved reliable in testing runs automatically.

What we clarify first

We clarify the task, examples from your daily work, the systems involved, how confidential the data is and who checks the results.

Our approach

Further reading

More on the topic.