Technology for context-based AI in Switzerland
SteP-met connects existing learning, business and data systems with context-based AI. So rather than generating information in isolation, the technology uses relevant, authorised data from a specific application context. Secure interfaces connect existing systems, context preparation, AI analysis, and practical support.
From existing systems to practical AI support
SteP-met’s core technological principle follows a clear data flow. It starts with existing learning, performance, support and business data. This data is accessed through a suitable interface and then prepared to underpin the respective usage strategy.
Data & systems → secure integration → context → AI → targeted support
The basic principle follows a clear process
Data source
Existing data forms the core element. Depending on the area of application, this may include learning tasks and learning progress, training and performance data, or information from knowledge, support and business systems.
Secure interface
A suitable integration connects the data source to the SteP-met system. Depending on the existing infrastructure, this may involve LTI 1.3, LMS integrations, APIs, data providers or custom interfaces.
Context preparation
Not every piece of information available needs to become part of an AI analysis. SteP-met selectively provides the relevant, authorised information required for the defined usage strategy.
Context-based AI analysis
The AI processes the context provided. The focus is therefore not on the broad, indiscriminate generation of new content, but on analysis based on the actual information contained in the respective system and the defined usage strategy.
Practical outcome
Depending on the area of application, the prepared context can give rise to analyses, recommendations, personalised learning support, training challenges or assistive functions. The outcome is always determined by the objective(s) defined in advance.
Technology-agnostic rather than tied to a platform
SteP-met is not tied to one single platform or interface. Existing learning platforms, enterprise systems and data sources can be accessed through appropriate standards, APIs or tailored integrations. This in turn makes it possible to transfer the underlying SteP-met principle to different technical environments and areas of application.
LTI 1.3 - LMS integration - APIs - data providers - custom interfaces
This open architecture is particularly relevant for companies and institutions that already use particular systems and data assets. Rather than fundamentally replacing existing structures, the focus is on assessing how these can be utilised to provide a secure context for AI-supported analyses and support functions.
Modular architecture, analytics and scalable operations
SteP-met's technological approach combines context-based AI with analytics and a modular system architecture. This allows solutions to be tailored precisely to a defined usage strategy and existing data to be used for suitable analysis or support functions.
To control and run operations, SteP-met uses Microsoft Azure as a scalable operating and deployment environment. Data protection, pseudonymisation and the protection of confidential implementation details are then handled separately as central areas of trust and security. Each specific technical implementation is tailored to the requirements of the respective company or institution.
Headless middleware
SteP-met is designed as headless middleware – an intermediary layer without its own user interface that sits between existing systems and context-based AI. This makes it possible to extend learning, training or business systems in a targeted way without replacing anything else. Access is provided through secure interfaces such as LTI 1.3, APIs or custom integrations, via which relevant data is then made available in a controlled manner. Existing investments in learning platforms, support- and business systems are thus preserved and selectively enhanced with context-based AI functions. Each specific connection and integration depends on the existing system landscape of the respective company or institution.
Explore the technology used in your specific context