Secure AI Chat and Flexible Assistant Framework for Enterprise Data
TellusR Dialogue is a scalable microservice that provides secure AI-driven chat against your business's selected content, whether internal or external. The solution is delivered as a standard product and can be installed behind your own firewall for maximum data control.
Retrieval
Utilizes TellusR AI search as the retrieval component to locate and rank relevant content across selected sources. Boost graph: Dynamic boosting/filtering of content using LLM. Increases precision in selecting relevant context by leveraging knowledge about where the chunks are located within the content structure or are categorized.
Assistant Configuration
Support for building custom assistants and assistant pipelines.
Utilize pre-configured assistants with proven standard setups. Your first chat assistant against your own data is just a few clicks away.
Examples of other predefined assistants:
Assistant for automatic generation of metadata based on a given document (e.g., summary, title, category, etc.)
Document-level chat, where users can ask questions about specified individual documents and receive answers based on them.
AI overview as an informative, relevant, and traceable text supplement to a search result (AI search +).
Support for MCP (Model Context Protocol).
Built-in prompt editor.
All functionality exposed in APIs.
Observability and Quality Assurance
Includes testing frameworks with support for both static tests and LLM-based assessments, allowing you to validate changes, compare versions, and detect regressions before deployment.
Full traceability and in-depth debugging throughout the chat pipeline, with insights into context, prompts, intermediate steps, and generated responses – significantly simplifying troubleshooting and quality assurance.
“The DiBK's internal work tool, Arkivslurpen, is built on TellusR and reviews the several thousand inquiries the directorate receives each year, extracting information from previous responses to make case processing in our response service faster and more accurate. This enables less experienced employees to respond both quickly and accurately, while ensuring full control over the data sources.”
– Kari Befring Bjørnstad, Director of Innovation and Digitalization at the Directorate for Building Quality
Chat with selected, pre-approved knowledge – never openly connected to the internet.
All processing and storage occurs on your terms, whether locally or in your chosen cloud.
Freedom to choose a large language model – avoid vendor lock-in and easily switch between different cloud services or use your own models, with low switching cost.
Install anywhere, seamlessly integrated into existing IT architecture – typically production-ready in hours, not weeks.
Built-in AI testing framework that ensures high production quality and reliability through continuous testing and validation.
Logging, traceability, transparency, and auditing provide documented control over interaction and data flow.
The dialogue only uses the dataset you define – perfect for handling sensitive content, internal guidelines, or specific professional systems.
Significantly faster information gathering with AI chat than through traditional searches or support systems.
Reduces errors and data leaks compared to open AI services – ensuring that sensitive information never leaves the company.
Increased employee productivity: Instant access to updated, quality-assured information – resulting in fewer interruptions during the workday.
Modern user experience: The same functionality as familiar AI chats, but with your security premises – and no compromise on privacy.
Invest in insight, security, and flexibility – not risk. Send us your most challenging use case!
