Back to list

News

19/12/2024

INTERGATOR 6.6

Enhanced Features for Semantic Search and Efficient Information Processing

With the release of INTERGATOR 6.6, the enterprise search solution has been significantly enhanced with features tailored to the needs of modern knowledge management systems. By integrating text embeddings, independent LLMs (Large Language Models), and a dialogue-based chat app, information retrieval and processing become more efficient and targeted.

Text Embeddings

Context-Based Optimization of Search Results

The introduction of text embeddings in INTERGATOR 6.6 represents a significant improvement in the semantic analysis of search queries. Unlike earlier word embeddings, which only analyzed individual terms in isolation, text embeddings transform entire queries into vectors and analyze them holistically.

Technical Advantages

  • Comprehensive Query Analysis: The entire context of the query is captured rather than isolated keywords.
  • More Accurate Results: The search engine delivers results that semantically align with the query.
  • Improved Relevance: Search results account for contextual relationships and similar meanings.

This approach reduces irrelevant results and enhances the quality of hits by efficiently interpreting natural language queries.

Independent LLMs

Control Over Model Operation and Data Sovereignty

INTERGATOR 6.6 enables the operation of Large Language Models (LLMs) within an organization’s own infrastructure. This solution is particularly relevant for companies with high requirements for data protection and sovereignty.

Key Advantages

  • Data Security: All information and processes remain within the organization’s boundaries.
  • Independence: No reliance on external APIs or providers like Azure or OpenAI.
  • High Availability: Running a local LLM ensures continuous availability without external latencies.

Operating LLMs requires a suitable GPU infrastructure but offers long-term benefits through full control over the systems and cost efficiency at high scalability.

Chat App

Efficient Answers Instead of Long Result Lists

Another highlight of INTERGATOR 6.6 is the integrated chat app, which serves as a dialogue-based tool for information retrieval. While traditional searches provide result lists, the chat app offers direct answers to natural language questions.

Chat-App in INTERGATOR 6.6
Chat-App in INTERGATOR 6.6

Chat App Features

  • Processes natural language questions to retrieve information.
  • Delivers clear and precise answers from the company’s knowledge base.
  • Source citations for transparency and traceable results.
  • Additional functionalities such as summarizing relevant documents or answering specific follow-up questions.

The chat app complements traditional search and integrates seamlessly into existing workflows. It provides an interactive solution for efficiently leveraging extensive data repositories.

The release of INTERGATOR 6.6 introduces significant advancements in semantic search and dialogue-driven information retrieval. Text embeddings analyze search queries in their entirety, while independent LLMs offer data sovereignty and flexibility. Augmented by the chat app, the search experience is enhanced with a dialogue-based component that delivers targeted answers and optimizes workflows.

INTERGATOR 6.6 supports companies in accessing their information faster, more efficiently, and securely.

Further articles