November 3, 2025

The November release of our RheinInsights Retrieval Suite brings improved and configurable query and AI pipelines. Now, our customers and partners can more easily configure and build complex workflows which leverage precise knowledge retrieval in the backend.

Use Cases

Our new query pipelines make it possible to easily summarize documents, data and knowledge on for specific topics, time frames and always also from disparate sources. It becomes particularly easy to answer support questions or write comprehensive how-tos based on knowledge which is scattered on multiple pages. All AI powered but with full understanding of the permission schemas in the underlying knowledge sources. No knowledge will be exposed or overshared.

Multilingual Linguistics

In order to allow for excellent search results, our suite implements not only vector search but also hybrid (semantic) search. Hybrid search combines the advantages of keyword search queries with those of vector search, resulting in significantly better (more relevant) search results.

However, the quality of keyword queries varies considerably depending on the linguistics used. If lemmatization or stemming is poorly chosen, the search results will be less good than with well chosen linguistics.

Therefore, the November release introduces new indexing schemes for the supported open-source search engines Elasticsearch, OpenSearch, Apache Solr, and for Azure AI Search. The new schemas are meant to support all common languages and can be easily extended towards your use cases.

Getting Started

After upgrading to our RheinInsights Retrieval Suite Version 1.12, the query pipelines can be straight-forwardly used through our query APIs, with our Slack and Teams integration and also with our stand-alone search interface.