INTERGATOR 6.8
Research organizational knowledge more flexibly and use it with confidence.
With Agentic Chat, enhanced hybrid search, and directly linked sources, INTERGATOR helps users address complex information needs more precisely, assess results more quickly, and complete recurring tasks more efficiently.
- Context-aware research with Agentic Chat
- Hybrid search for short terms and natural-language queries
- Traceable answers with inline references and document sources
Organizational knowledge is available
… but research and interpretation take time.
Important information is often spread across documents, file repositories, business applications, and other connected systems. The knowledge exists, but when it is needed, it still has to be found, reviewed, and placed in the right context.
Short search terms often return too many results. Complex questions require several search attempts, while relevant documents have to be opened and compared individually. AI-generated answers are only truly useful when users can see which organizational sources they are based on.
This is exactly where INTERGATOR 6.8 comes in: with improved search, a more flexible chat experience, and traceable source references.
Agentic Chat
Research more flexibly and respond more precisely.
With INTERGATOR 6.8, the chat is fundamentally enhanced. Instead of automatically starting the same research process for every input, the language model first analyzes the user’s intent and decides on the next steps based on the context.
Depending on the request, Agentic Chat can:
- respond directly when no research is required,
- search INTERGATOR for relevant organizational information,
- ask targeted follow-up questions when a request is unclear,
- perform multiple searches and evaluate the results together, depending on the model used.
This makes it possible to work through more complex questions step by step. The chat can adapt its search queries to the course of the conversation and does not have to treat every follow-up question as an entirely new research task.
Chat scopes remain available. Users can, for example, limit the chat to selected documents, favorites, or defined document collections. However, this restriction is no longer generally required before starting a conversation.
Agentic Chat does not operate fully autonomously. In INTERGATOR 6.8, it primarily has access to INTERGATOR Search as a tool. Which capabilities can be used and how extensive the research can be depend, among other factors, on the language model in use and the respective configuration.
Hybrid Search
Better support for short search terms and complex queries
INTERGATOR 6.8 brings lexical and semantic search methods even closer together. Instead of having to choose a single search approach, it combines the strengths of both methods.
As a result, the search supports different research habits:
- individual terms, identifiers, or numbers,
- short and precise search queries,
- complete questions in natural language,
- detailed descriptions of information needs,
- multilingual search queries.
Lexical search considers exact terms and linguistic variations. Semantic search also recognizes contextual relationships and similar phrasing. INTERGATOR combines the results of both methods and prioritizes the most relevant matches overall.
Users can therefore continue to search for a known project name, an invoice number, or an individual technical term. At the same time, they can formulate complex questions without having to know the exact wording used in the relevant documents.
INTERGATOR also uses a language model to identify relevant terms and passages within the search results. This makes it easier to understand why a document matches the query, even for natural-language or multilingual searches.
KI-tasks
Complete recurring tasks more efficiently
INTERGATOR 6.8 makes it easier to access frequently used AI tasks. Users can pin important tasks and then launch them directly from the search results or the document preview. Freely formulated questions and instructions can also be saved and reused. This means recurring reviews, summaries, or analyses no longer need to be entered from scratch each time.
This is particularly useful for recurring workflows, for example:
- review documents against defined criteria,
- summarize key content,
- identify main topics or relevant statements,
- apply the same questions to multiple documents,
- reuse prepared instructions for further research.
What is an agentic chat
Agentic Chat no longer follows the same rigid process for every request. The language model can determine whether to respond directly, search organizational knowledge, or ask a clarifying question first. This article explains how it works, its benefits and limitations, and how it is implemented in INTERGATOR 6.8.
Discover how INTERGATOR 6.8 changes the way your organization works
In a personalized demo or one of our release webinars, we will show you Agentic Chat, enhanced hybrid search, and the new options for recurring AI tasks—using practical examples tailored to your requirements.