AI in everyday work
AI in trades and SMEs: three workflows that give office time back
Quotes, work reports, looking up regulations: Swiss businesses lose their evenings in the office, not on the job site. This article walks through three workflows. You see how an AI colleague works with your company’s knowledge, what can connect to your software, and where the limits are.
In brief
- Translation (52%) and correspondence (47%) are the most common AI applications in Swiss SMEs, exactly the work that piles up in the evening.3
- Kisuno works with your company’s knowledge: quote templates, grid operators’ utility regulations, work instructions and past reports become a knowledge base you can query.
- Your data stays in Switzerland or on your own server and does not train any models.
- Software with an open interface can be connected, for example Microsoft 365, bexio or KLARA. An interface is the point where two programs exchange data. The route for this is called the Model Context Protocol (MCP): with permissions per tool and a log entry for every access.
Starting point: administration grows faster than the business
Around 60% of Swiss companies perceive their administrative burden as high or fairly high. Construction projects weigh heaviest, at 67%.1 Counted across the whole country, avoidable bureaucracy adds up to around 30 billion francs per year, according to a study by ifo and BSS.2
In the trades the direction is clear: in a German association survey of 10,630 businesses, 74% report that bureaucratic effort has grown over the past five years.4 In the electrical trade there is deadline pressure on top. The grid operator asks the property owner in writing to submit the safety verification report (SiNa), at least six months before the inspection period expires, that is the interval up to the next periodic inspection. If the report is still missing after two reminders, enforcement passes to ESTI.5 The deadline pressure lands on the business that repairs the installation and delivers the paperwork.
At the same time, AI has arrived in SMEs: within one year, the share of Swiss SMEs using AI rose from 22 to 34%, most often for translation and correspondence.3 So the question is no longer whether a business uses AI. The question is whether it works with the company’s knowledge, and where the data ends up.
- Keywords from the job site, typed or dictated
- Draft built from templates and company knowledge
- Review and approval in the business
- Sending and filing, fully logged
Use case 1: quotes and correspondence from your own company knowledge
An enquiry arrives in the morning by email: bathroom renovation, three keywords, two photos. The AI colleague knows the company’s quote templates, price lists and past quotes. It suggests line items, writes the cover text in the company’s tone and files the draft for review. You remove one item, adjust one price and send the quote.
The same applies to the correspondence around it: rescheduling, payment reminders, replying to a complaint, a translation for a French-speaking client. The draft is built from the previous correspondence and your own documents; approval stays with a person.
- The customer’s enquiry, as brief as it arrives
- The company’s templates, price lists and past quotes
- A quote draft in the company’s tone, ready for review
The difference to a chat in the browser: the draft is based on your documents rather than on pasted-in snippets. And your quote and costing stay in your own instance instead of in someone else’s chat account. An instance is your own copy of the system, kept separate from other customers.
Use case 2: from site notes to a clean report
On the way home the technician types or dictates three sentences into the phone; the device’s dictation function is enough. He notes what was done, what was used and what stood out. Out of that come the time-and-materials report in the company format, the materials list and a defect description for the site management: worded, numbered, ready to file. A summary of the quantities carried out, the site measurement, is drafted too. The checking stays with the business.
In the electrical trade, clean paperwork is about more than tidiness: measurement and test protocols and the safety verification report follow fixed forms.6 Filling in and checking them remains the job of the authorised inspector. The AI colleague takes over the writing around them, from the cover letter to the grid operator to the defect list for the customer.
- Three short sentences from the drive home
- One input, three documents in the company format
- Report, materials list and defect description, ready for approval
What deliberately does not happen: no document is sent automatically. Every report waits for approval, and every change remains traceable.
Use case 3: the company’s knowledge becomes queryable
Which utility regulation applies to the metering installation with this grid operator? How did we service this installation two years ago? Today, questions like these interrupt the management. A knowledge base is a collection of your own documents that the AI colleague is allowed to search: work instructions, utility regulations, manufacturer documents and past reports. From those it answers such questions and names the source. That way the answer can be checked in seconds.
New employees ask in their own language and get answers from the same documents. The safety briefing in Portuguese is generated from the company’s German fact sheet. Important for standards: texts such as the Swiss low-voltage installation standard (NIN) are copyrighted. Whether they may go into a knowledge base depends on your Electrosuisse licence. Clarify the scope before uploading. Your own documents belong there anyway.
- A question in everyday language, on the phone or in the office
- Your own documents, with permissions per tool and a full log
- An answer with its source, checkable in seconds
Permissions apply down to individual tools: whoever may not see the payroll folder does not see it through the AI either. Every access is recorded in the audit log, the running record of who retrieved what and when.
Connectors: what can be linked to your software
An AI colleague is only as useful as what it is allowed to reach. A connector is the ready-made link between Kisuno and another program. It becomes possible when that program has a documented API, that is, an interface through which programs exchange data. The framework for this is MCP, an open standard for connecting external systems. Permissions apply per individual tool, and every call is logged. This is the state of play for common trades and SME software, checked in August 2026, from the AI connection to bexio and KLARA through to Microsoft 365:
| Software | Interface | Connection to Kisuno |
|---|---|---|
| Microsoft 365 (Outlook, Teams, SharePoint) | Graph API, openly documented | MCP server available from Microsoft; email, calendar and files directly addressable |
| bexio | REST API, openly documented | Connection via MCP possible: contacts, quotes, orders, invoices, projects |
| KLARA | REST API, openly documented | Connection via MCP possible: accounting, sales, payroll |
| CashCtrl | REST API, openly documented | Connection via MCP possible: accounts, orders, reports |
| Odoo | Open interface, documented | Ready-made MCP connections available from the community |
| Abacus | REST API with secured sign-in (OAuth2), access through the partner programme | Connection via MCP possible; access runs through the Abacus partner channel |
| SORBA, Messerli, ElektroForm | No publicly documented API | The route runs via exports, via the standardised construction data exchange, or via the connected accounting software. The basis is the SIA 451 standard; for bills of quantities the IfA18 profile with the CRBX file format has applied since 20207 |
Is your software missing from the list? Much of the daily work runs without any system connection at all, from quotes to the knowledge base. How a connection via MCP works in principle, and which systems are prepared, is described on the integrations page.
Data protection: your costing stays in the business
Quotes, costings and customer files are trade secrets. Under the revised Swiss Data Protection Act, the business remains responsible for personal data even when an AI tool processes it. If a provider processes that data for you, a data processing agreement is required. That is the contract setting out what the provider may do with the data. This article is not legal advice.
Kisuno is operated in Switzerland or runs on your own server. Customer data does not train any models. On offer are the current models from OpenAI (ChatGPT) and Anthropic (Claude); they run in the AWS data centre in Frankfurt. AWS is a US company, so the US CLOUD Act applies even with servers in Frankfurt, and we say so openly. For customers with higher data-protection requirements, open frontier models such as GLM, Kimi and DeepSeek are available, operated by European companies on servers in Europe; the CLOUD Act does not apply there. The data processing agreement and the subprocessor list are published openly on the website.
In practice this means: the price list you calculate with ends up in nobody’s training data, and when a customer asks where the data is held, you answer in one sentence.
Frequently asked questions from trades and SMEs
My industry software has no interface, for example SORBA or ElektroForm. Is Kisuno still useful for me?
Yes. Quotes, reports, correspondence and the knowledge base work with documents and need no system connection. Where a connection is missing, the route runs via exports, via the standardised construction data exchange (the SIA 451 standard, for bills of quantities in the IfA18 profile since 2020), or via the connected accounting software.
Does the AI write my quote on its own?
No, it writes the draft from your templates and price lists. Line items, prices and the approval stay with you. A language model can produce errors. That is why every text is a draft for review, not a verified statement.
May I load the NIN standards into the knowledge base?
Standard texts such as the NIN are copyrighted. Whether you may load them into the knowledge base depends on your licence from Electrosuisse. Clarify this before uploading. Your own work instructions, the grid operators’ utility regulations, manufacturer documents and your reports make up the larger part of your company knowledge, and they are yours anyway. This answer is not legal advice.
What does it cost for a ten-person business?
The managed instance costs CHF 7 (EUR 7.50) per person per month. Through the end of the current month: 50 percent off with a twelve-month term, that is CHF 3.50 (EUR 3.75) per person per month. Testing is free and not limited in time: every new account starts with a starting credit of CHF 50 (EUR 55) for model usage, with no payment details. If you self-host, the software costs nothing, permanently and from a single user. The LibreChat base is available under the MIT licence; the repository with our own extensions follows shortly. All details are on the pricing page.
Where is our data held?
In the managed variant in Switzerland, with self-hosting on your own server. Customer data does not train any models. The models from OpenAI (ChatGPT) and Anthropic (Claude) run in the AWS data centre in Frankfurt; AWS is a US company, so the CLOUD Act applies there too, and we say so openly. For higher data-protection requirements, open frontier models such as GLM, Kimi and DeepSeek are available from European providers on servers in Europe; the CLOUD Act does not apply there. The evidence is available in the Trust Center.
Do we need an IT department for this?
No. In the managed variant the operator takes care of operation, updates and backups. You manage users and permissions in the admin panel, your organisation’s administration screen. For self-hosting you need someone who can maintain a Linux server with Docker, which can also be an IT service provider.
Does it work on the phone, on site?
Yes. Kisuno runs as a PWA on iPhone, Android devices and the desktop. A PWA is a web application that can be installed on the device like an app. Taking notes, photographing and looking things up works where the work happens; your device’s dictation function is enough. The finished report waits in the office for approval.
See it on a case of your own
Bring a quote, a work report or a fact sheet, and we will walk through the workflow together. Or sign up: because demand is high, we hand out access to the managed Swiss instance in stages.
Sources
All figures in this article come from the following sources, last checked in August 2026. The German association survey is marked as a Germany figure.
- SECO Bürokratiemonitor (bureaucracy monitor, survey 2022, published 2023), perceived regulatory burden: seco.admin.ch
- ifo Institute and BSS Volkswirtschaftliche Beratung for economiesuisse and other associations (November 2025), costs of avoidable bureaucracy: economiesuisse.ch
- AXA SME labour market study 2025 (surveyed in March 2025, published on 8 October 2025, 300 SMEs), AI use in Swiss SMEs: axa.ch
- ZDH, special survey on bureaucratic burden in the trades (survey 2023, Germany, 10,630 businesses): zdh.de
- ESTI, low-voltage installations: periodic inspection and the grid operator’s written request under Art. 36 NIV, safety verification report under Art. 37 NIV: esti.admin.ch
- EIT.swiss, together with Electrosuisse, ESTI, VSEK and VSE, form for the safety verification report on electrical installations including the measurement and test protocol (available to members): eit.swiss
- CRB and SIA, standard SIA 451 as the framework for data exchange in construction; for bills of quantities the CRB directive IfA18 with the CRBX format has applied since 1 January 2020: crb.ch
Last reviewed: