And scales all the way to specialist agents on Microsoft Foundry.
Dinotronic brings AI into everyday work without you losing control over data, permissions and compliance. From your first Copilot task to your own agent architecture, supported from Zurich.
Which starting point is right for you?
Copilot summarizes emails, takes meeting minutes and answers questions about your documents. Getting started stays simple, but not uncontrolled.
Recommended starting pointCopilot Quick Start
Discuss your startVery few companies start with an AI strategy. They start with a full inbox, long meetings and scattered knowledge. That is exactly where AI should help, and exactly where it needs guardrails.
Summarizing the inbox, writing up meeting notes, understanding documents faster, preparing quotes. These are not big projects, but everyday tasks that save time from day one.
If Copilot works with permissions that have grown over the years, or employees use unapproved tools, data protection, reputational and compliance risks arise that nobody notices until it is too late.
Specialist agents need architecture, data quality, access control, monitoring and someone who still understands and runs the environment after the project.
Start simple
You do not need a strategy, you need less effort. This works from day one.
Go further
You have the basics in place and want to apply AI to your processes and data, in a controlled and verifiable way.
You do not have to start with a big project. Each offering works on its own, and each one prepares the next.
Half a day in which Copilot is switched on in your environment and your team is trained on it: emails, meetings, documents and research, practiced on your own examples. Followed by a short look at actual usage. For MWS Mega customers the feature is already included, and this adds the hands-on guidance.
We collect use cases, assess benefit, effort, data situation and risk, and turn them into a roadmap for the first 30 to 90 days, written for executive management.
AI inventory, unapproved usage, gap analysis against FINMA Guidance 08/2024 and the revFADP, a signature-ready AI policy and a management report. Three weeks, around six hours of effort for your team.
Permissions, SharePoint structure, data classification, Microsoft Purview, DLP and Conditional Access. The foundation that ensures AI only accesses what it is allowed to see.
Your own agents for business processes: internal knowledge search, document pre-screening, minutes, quote comparisons. With Copilot Studio without code, or with Microsoft Foundry including integration with CRM, ERP and DMS.
Copilot helps quickly, agents achieve a lot. Both need data quality, clean permissions and clear accountability. That is why we see AI not as a project, but as part of the environment we already run for you.
Everyday work
Outlook, Teams, SharePoint, Word, Excel. This is where AI delivers value first, because that is where the work already happens.
Governance
AI inventory, roles, policy, data classes, DLP and monitoring. This makes AI manageable, instead of banning it or letting it grow unchecked.
Copilot and agents
Copilot Studio, knowledge agents, automations. Agents get clear boundaries: defined data sources, permissions and approvals.
Foundry and integration
Microsoft Foundry, Azure AI, interfaces to CRM, ERP and DMS. This is where a demo becomes an operation that someone monitors and keeps developing.
Every industry has different documents, different deadlines and different guardrails. The starting point stays the same: your processes, your data, your risk. Choose what you work with.
The starting point is usually pragmatic: less email stress, better follow-up after meetings, faster access to internal knowledge and the first clear rules.
Understand long threads, draft replies, spot priorities faster.
Derive decisions, tasks and next steps from Teams meetings.
Make recurring content, client data and templates faster to use.
Before wider use, make sure Copilot does not see more than intended.
What is Copilot allowed to do? When do you need governance? And when do your own agents pay off?
AI consulting helps you find worthwhile use cases, roll out Microsoft 365 Copilot securely, prepare your data and permissions and, where it makes sense, build your own specialist agents.
In Switzerland, the revised Federal Act on Data Protection (revFADP) applies, and depending on your industry there are supervisory expectations as well, such as FINMA Guidance 08/2024.
Copilot summarizes emails and Teams meetings, answers questions about Word documents and Excel spreadsheets, drafts replies and minutes, and speeds up research.
The benefit depends on the role, not the size of the company. People who write, read and take minutes a lot save the most time. People who spend their day in a specialist application save far less.
It shows which AI tools are actually in use, which data is affected, what risks exist and which rules, roles and technical measures you need.
With us it takes three weeks, starts at a package price of CHF 8,900 and delivers an AI inventory, a gap analysis, a signature-ready AI policy and a management report. Details of the check.
That depends on your tenant and the services you have booked. Microsoft offers data residency in Switzerland for Microsoft 365, but some services and processing steps run outside the country. We check and document what applies to your environment before rollout, rather than making blanket promises.
For Copilot, your content stays within your tenant and is not used to train the models. Permissions matter at least as much. Without a clean structure, Copilot will also find documents an employee should never have seen.
As soon as agents need to connect to specialist systems, work with your own knowledge sources, or be monitored and versioned.
For agents built on approved SharePoint and Teams content, Copilot Studio is enough, and your subject matter experts do not need any programming skills.
Yes. Small organizations benefit first from simple tasks: summarizing emails, taking meeting minutes, analyzing documents, preparing quotes.
What matters is a small start with secured permissions. We do not sell licenses on speculation. We start with a pilot group and measure actual usage.
Yes. Copilot summarizes individual emails, long threads and whole sections of your inbox, highlights open items and suggests replies. The same works for Teams meetings: Copilot produces minutes, decisions and action items.
You need a Microsoft 365 license with Copilot and a cloud mailbox. We set it up and show your team, using real examples, how to phrase requests that get good results.
Copilot is the finished product inside your Microsoft 365 apps. It works with the data each employee already has access to and can be extended with your own agents in Copilot Studio, without programming.
Azure AI Foundry is the development platform behind it. That is where you choose models, connect your own data sources and specialist systems, control costs and monitor operations. In short: you use Copilot, you build with Foundry.
Yes, as soon as employees use AI tools. The revFADP requires you to know and document how personal data is processed. Without a rule, every person decides for themselves what to enter into a tool.
An AI policy for an SME does not have to be long. Eight to twelve pages that clarify responsibilities, approved tools and the handling of client data are enough in most cases and can be ready in three weeks.
In three steps. First, find out what is already being used, usually more than management expects. Second, review permissions and file storage so AI cannot access documents it should not see. Third, start with a pilot group on real tasks and measure usage.
What we do not recommend: buying licenses for everyone and hoping the benefit will appear on its own. That is the most common reason AI budgets fizzle out without impact.
For workshops and readiness assessments, market prices typically range from CHF 2,000 to 5,000, with day rates of CHF 1,200 to 2,500. These formats provide orientation, but no documents you can put into practice.
Our AI Governance Check starts at a package price of CHF 8,900 and delivers an AI inventory, a gap analysis, a signature-ready policy and a management report. You receive the binding fixed price after the free initial consultation.
That is exactly what we are here for. You take on neither project management nor technology, you bring your knowledge of the business.
For the AI Governance Check, that means around six hours of your time, spread over three weeks.
Find your starting pointWhich step fits you?
Get started