Executive Briefing · Research and knowledge providers · 26 November 2026

AI for research and specialist knowledge. Your data remains your asset.

Unpublished results, research data and licensed content are the capital of institutes and knowledge providers. In this Executive Briefing, we show how AI agents work with them without your content ever leaving your environment.

26 November 2026
11:00 · 60 minutes
Online and free of charge
Sovereign AI OS von in-manas: Operations-Dashboard mit Ausführungen, Fehlerquoten und Warnungen der KI-Agenten je Abteilung
in-manas is deployed wherever data needs special protection.
EURAC researchEAMPapstarJUSTBoschFFGUniversität InnsbruckBiolyz

Patient data

Critical infrastructure

Strategic company data

Research

Where AI accelerates research and specialist knowledge, intellectual property is at stake.

Literature and source research

Agents search publications and internal holdings and summarise the current state of knowledge.

Critical: search queries reveal what you are currently working on.

Analysing research data

Agents prepare measurement series, study data and interim results.

Critical: unpublished results and personal study data.

Specialist information for customers

Knowledge providers make their content available to subscribers as an AI application.

Critical: licensed content may only be seen by authorised users.

Proposals and reports

Agents draft grant proposals and project reports from existing documents.

Critical: ideas and budgets are confidential until submission.

AI risks and answers

The risks are well known. The AI OS has an answer for each of them.

Risk
How the AI OS responds
Knowledge reaches third parties.

Unpublished results and content are processed by external services.

Operated in Europe or on your premises.

The AI OS runs in your data centre or in a European cloud. Your data never leaves this environment.

Unclear access.

Nobody knows exactly which application accesses which holdings.

Permissions per agent.

A manifest defines what an agent is allowed to access. Denied access attempts are visible.

Results without provenance.

Answers cannot be traced back to their sources.

Every step traceable.

For every agent run, the data and tools it used are documented.

Isolated solutions per working group.

Every team builds its own tools, and nothing can be reused.

One foundation for everyone.

Agents are built with the SDK and CLI on a shared platform and can be shared across teams.

You stay in control at four key points.

Sovereignty is built into the AI OS from the very first installation, at no extra cost.

Contractual partner

A European company headquartered in Austria.

Infrastructure

Your own data centre or a European cloud, with the same functionality.

Model execution

You decide where models run: locally or with a provider of your choice.

Control and logs

Permissions, logs and costs remain in your hands.

Executive Briefing · 60 minutes · online

AI in research and specialist information: using knowledge without giving it away

26 November 2026
11:00 · 60 minutes
Online and free of charge

For leadership and IT teams at research institutions, publishers and specialist information providers.

Agenda
1
Where AI stands today in research and specialist information
2
Critical workflows and their risks
3
Live demo: a research agent with permissions and traceable results
4
Your questions
Speakers
Stefan Unterhiner
Stefan Unterthiner
[Role], in-manas
Dieter Tschemernjak
Dieter Tschemernjak
[Role], EURAC or research partner

Registration

Secure your place

You will receive the access link by email. Everyone who registers will receive the recording afterwards.

Thank you for registering! You will receive the access link by email shortly. Please also check your spam folder.
Unfortunately, something went wrong. Please check your details and try again.

Can't make this date? Book an initial consultation

Frequently asked questions

How much does the Executive Briefing cost?

Nothing. Participation is free of charge.

Will I get a recording?

Yes. Everyone who registers receives the recording by email, even if they could not attend live.

Can we offer our content as our own AI application?

Yes. The AI OS does not come with a fixed user interface. You connect your own application and keep content, permissions and logs in one place.

Can our own developers build agents?

Yes. Agents are built with the SDK and CLI in your team's language and live in your repository.

Which models can we use?

Local models as well as models from providers of your choice. Switching is simply a configuration change.

Register now