All training programmes — Winterthur Contact us

AI training in Winterthur

In Winterthur, AI training starts from technical documents and maintenance sheets. Participants learn to rewrite without changing units, warnings or the order of operations, and to spot what a tool has added without evidence.

Start with real work and a local context

The context below is a fictional teaching example, not a client testimonial. Scoping selects an exercise relevant to your work: drafting, handovers, document analysis or decision preparation. Quality is assessed against reference material and criteria agreed before the exercise.

Rewrite a fictional maintenance sheet without changing units, warnings or the order of operations.

Which AI training path fits your needs?

AI on the Job — Winterthur
Identify a real task, write an instruction and compare the output with reference work.
Managing in the AI Era — Winterthur
Decide what can be assisted, who checks quality and when an error must be escalated.
Leading AI Adoption in Teams — Winterthur
Select a pilot use case, define evaluation criteria and organise team feedback.
Building No-Code AI Assistants and Agents — Winterthur
Scope an assistant, restrict its sources and permissions and test its behaviour before deployment.

Claude AI training in Winterthur

For Claude AI training in Winterthur, Claude, Anthropic’s assistant, can support a workshop when your organisation permits its use. The aim is not brand promotion: participants learn to scope a task, supply sufficient context and examine the result. Tool selection and features are confirmed during scoping according to available access.

An exercise with Claude can use a fictional document pack: request a summary restricted to the supplied sources, separate facts from interpretations and trace every assertion back to the documents. Participants compare an initial prompt with one specifying the audience, constraints and output format. Better wording is accepted only when the answer also respects the facts and limits of the pack.

Write and refine an instruction for a workplace task.
Analyse an authorised document and verify the passages used in the answer.
Compare responses, detect omissions and correct unsupported claims.
Define human approval before reusing an output at work.

Claude is an Anthropic trademark. This page claims neither an Anthropic partnership nor an official Claude certification. No Claude subscription is advertised as included.

A learning process built around verification

Start with a work example and define what an acceptable answer would look like.
Test using fictional or anonymised data and retain outputs for comparison.
Check facts, omissions and errors; document human approval.
Define what can transfer to the workplace, what remains prohibited and what needs review.

Data, errors and human supervision

Before testing a tool, define permitted data, reference sources and who approves outputs. Exercises use fictional or anonymised data. Fluent answers can be wrong; a no-code prototype is not a production-ready system.

Prepare your AI training

Specify the roles involved, participants’ experience, permitted tools and a representative task. Do not send confidential documents with your enquiry. Delivery arrangements, working language and venue are agreed with SHR; a service city is not an SHR office.