How Enterprise Search works

We start by indexing all files and documents from the data sources to be searched. This index is searched for matching results during a search query. A sorting of all hits is performed according to differently adjustable criteria.

Already at the beginning of the indexing process, the access rights for the users are stored (early binding) and automatically taken into account in the user interface for each information search.

Our enterprise search software enables you to search across your company and thus access your most diverse applications and systems. This starts with your email systems such as Office365, in internal company drives, in wikis such as Confluence, or even in various file formats such as PDF, Excel and scanned documents, as well as video and audio files.

And this is how our Artificial Intelligence Suite is built:

Architecture Enterprise Search Platform

Optimal approach: AI and enterprise search

Through our approach, we combine enterprise search, artificial intelligence, machine learning with sophisticated linguistics.

This gives users centralized access to all structured and unstructured data. Likewise, the user can analyze content according to various criteria, e.g. personal data and set so-called “smart filters” to perform various ad-hoc analyses. Here you can choose whether the results are displayed via a classic hit list or a flexible dashboard.

The simplest manifestation is the search within an application, e.g. in the intranet, in Confluence (available in the Atlassian Marketplace or real-time search in Netapp storages with the world’s only NetApp-certified connector).

By connecting different data sources with our AI-based enterprise search approach, you get a centralized rights-checked search with a ready-to-use user interface.
This enables us, for example, to set up a Digital Workplace for your service staff in the service or help desk. The search time is significantly reduced and thus the total AHT (Average Handling Time).

If you already have an “ElasticSearch” system in use, we can significantly improve the search quality by using our ready-made Elasticsearch plugin for high quality linguistics and thesaurus search extensions….

Our enterprise search approach uses AI techniques (machine learning, transfer learning), high-quality linguistics for 28 languages, named entity recognition, tagging, text mining and text classification. Semantic search enables numerous sophisticated research use cases such as GDPR-specific analysis, content enrichment, and extensive data discovery and extraction scenarios, as well as case processing in the police environment.

Artificial Intelligence (AI) Suite for AI-based Search, Content Enrichment Usecases, NLP & NLU

Other elements offered by our enterprise search approach:

These include content enrichment, text mining, tagging, analytics with the AI suite, and ready-made products such as our “Document and Contract Analyzer”.

Content enrichment generates any active metadata and metadata layers, such as export control, confidentiality levels, personal data, deletion periods, and intellectual property.

We analyze any content for relevant data points, findings, clauses and text passages that are relevant for further processing.

We identify the required entities of a document, recognize the subject of a document, the document type and the text classification.

We deliver NLP-driven scenarios such as use in chatbots or analyzing your email communication with customers for automation in digital processes.

Furthermore, the analysis of e.g. legacy data to analyze customer satisfaction or cancellation reasons for you.

The most important functions of our solution

What is behind the buzzwords “Cognitive Search” and “Insight Engine” and what does it offer you?

Cognitive Search refers to the machine understanding of user intentions.
It uses artificial intelligence to better understand user behavior and through this we can provide the best possible answers to all search queries.

We deliver optimal search results through NLP, AI and high performance linguistics

  • Excellent relevance of the hit list

    You get search results that match your search query exactly

  • Semantic search

    The relationships are recognized across a wide variety of data silos

  • Aggregate data simply -

    in the hit list or via a Knowledge Graph

  • Depth analysis of the found documents possible -

    and thus significantly more than just a full text search

  • Ask search queries in natural language