How beatmet24 uses local AI
to support care staff and administrative teams
HEALTH & CARE SERVICESbeatmet24 uses the IOWIS local AI platform to support around 300 care staff as well as its office and administrative teams with recurring internal questions. The platform runs within beatmet24’s infrastructure. AI processing and the connected knowledge sources therefore remain within the company’s controlled environment. Care staff access the platform through Microsoft Teams during their day-to-day mobile work. Office and administrative teams use the AI platform’s central interface.
Photo: beatmet24 GmbH
300 care staff
plus office and administrative teams. Mobile staff and office-based teams use the same AI platform through the interface best suited to their work.
From staff rotas to employment contracts
When generating an answer, the AI only considers documents available to the individual user or their group.
No source, no answer
If the authorised material does not provide a reliable basis, the AI returns no substantive answer instead of generating a plausible-sounding but unsupported statement.
Making existing knowledge
easier to access
Photo: beatmet24 GmbH
Mandatory training and internal procedures also generate recurring questions whose answers are often already documented in existing materials
beatmet24 GmbH is a specialist home care provider for out-of-hospital intensive and ventilation care in the greater Stuttgart area and the Böblingen district. A large proportion of its care staff work away from the office. When questions arise, they cannot always access documents held there or reach the relevant points of contact at short notice. At the same time, office teams, administrative staff and management regularly receive similar questions. These concern matters such as mandatory training, internal responsibilities, terms of employment, emergency plans, technical problems or the use of MEDIFOX DAN care software.
In many cases, the required information had already been documented but was spread across different materials and systems. Staff therefore had to search for it manually or obtain it directly from another person each time. The actual bottleneck was not a lack of knowledge, but timely and targeted access to it.
Local AI for mobile care staff
and office teams
Photo: beatmet24 GmbH
Mobile care staff access the platform through Microsoft Teams and can ask questions there in natural language. Microsoft Teams provides the access channel for staff working away from the office; the AI and knowledge processing itself takes place within beatmet24’s infrastructure.
Office and administrative teams use the AI platform’s central interface. There, users can create general-purpose AI chats and their own workspaces, configure specialised AI assistants, add files and knowledge bases, and share chats with colleagues.
The system can therefore support both general AI tasks and questions that need to draw on internal company knowledge.
Knowledge sources and permissions
Photo: beatmet24 GmbH
Office and administrative teams require different internal information depending on their tasks. The platform only considers the knowledge sources available to the respective user.
For internal questions, the AI uses only the knowledge sources made available to the user. These include organisational documents, employee information materials, technical guidance and documentation for MEDIFOX DAN care software, as well as sensitive HR documents such as employees’ employment contracts.
A granular roles and permissions model applies the need-to-know principle. For every request, the platform only considers documents and knowledge available to the individual user or their group. Employees can, for example, ask about information in their own employment contract without gaining access to the contracts of other members of staff.
Questions about annual leave, overtime, notice periods or other terms of employment can therefore be addressed using the documents relevant to the respective user. Where a question involves employment law, relevant statutory provisions can also be considered alongside the internal documents.
Answers strictly grounded
in authorised sources
SOURCE GROUNDINGFor internal knowledge questions, beatmet24 uses a strictly source-grounded retrieval-augmented generation approach. The language model is configured to generate answers only from documents authorised for the respective user. If those documents do not provide a reliable basis, no substantive answer is returned. This configuration prevents plausible-sounding statements that are unsupported by the internal sources and reduces the hallucination risk associated with generative AI.
Local processing
of sensitive information
OPERATIONThe AI platform runs within beatmet24’s infrastructure. AI processing and the connected knowledge sources remain within the company’s controlled environment. This allows beatmet24 to use sensitive HR documents and confidential information from day-to-day operations while retaining control over how the information is processed and stored.
From technical readiness
to day-to-day use
IMPLEMENTATIONIOWIS is responsible for the complete technical implementation of the local AI platform at beatmet24. Once this work has been completed, the solution is ready for its intended use.
For organisational adoption, IOWIS works with its partner JADE IT. At beatmet24, JADE IT helps employees become familiar with the platform and embed its use in existing workflows.
Technical implementation by IOWIS
IOWIS provides the local AI platform and is responsible for its complete technical implementation at beatmet24. This includes setting up the platform and preparing it for its intended use.
Organisational adoption support with JADE IT
JADE IT supports organisational adoption within the company. At beatmet24, the focus is on helping employees become familiar with the platform, gathering feedback from day-to-day use and embedding new ways of working with the platform into existing processes.
Practical value and controlled AI use
With the local AI platform, beatmet24 addresses two requirements: resolving recurring internal questions more quickly and processing sensitive information in a controlled way. Care staff are less dependent on the availability of individual points of contact, while office teams, administrative staff and management receive fewer routine questions.
The key benefit is that AI use does not have to be limited to non-sensitive content. Personal and confidential documents can also be used because processing takes place locally and access rights are assigned selectively. This enables beatmet24 to use AI where it is most useful in day-to-day work.
What could a similar use case look like
in your organisation?
At beatmet24, the IOWIS Platform makes existing knowledge easier to use in day-to-day work and enables sensitive information to be processed within the company’s own infrastructure. In a personal demo, we show how this approach can be applied to your requirements.