With its latest release, Fabasoft Approve expands its AI capabilities and improves key document processes. The focus is on the new AI-supported metadata extraction, a smarter AI chat with additional metadata context, and improvements to search-based lists and the synchronization of requested documents.
AI-supported metadata extraction simplifies data capture
The new AI-supported metadata extraction automates the capture of information from documents. Based on freely definable models, the AI identifies key attributes and transfers them directly into the object’s corresponding metadata. This reduces the effort required for manual data entry and speeds up document processing.
The AI provides support right from the configuration stage: prompts, synonyms, and examples for the individual extraction definitions can be generated automatically. In addition, the AI analyzes not only text content but also tables, charts, technical drawings, and images. The feature is complemented by a JSON interface that allows metadata extractions and their definitions to be easily imported.
A new metadata extraction can be created in the configuration under the “AI Elements” list entry. Next, an extraction definition is created for each property to be extracted, and the corresponding target property is assigned.
Using the “Populate With AI” context menu, the AI automatically generates the appropriate prompt, relevant synonyms, and real-world examples for the property value based on the target property.
Once metadata extraction has been configured and approved for an object, it can be selected and started via “Extract Metadata.”
AI Chat provides answers with even more context
AI Chat now offers answers with even more context. In addition to the actual document content, the AI also takes into account the metadata of the current object. This includes, for example, status information, project assignments, responsibilities, or properties from categories.
Thanks to this additional context, the chat provides answers that are even better tailored to the respective object. At the same time, a redesigned layout ensures a clear user interface when chatting with structured objects, for which only the chat will be displayed in the future.
The split view with document preview remains active when chatting with individual documents.
Search-based lists are now even clearer
Search-based lists have also been optimized. Deleted items will no longer appear in search results or list views. Instead, they will be available exclusively in the trash of the respective room or via search folders.
As a result, search results focus on active content, enabling faster and more organized work.
Improved synchronization of requested documents
Synchronization between documents and requested documents (document instances) has also been enhanced. In addition to metadata, Fabasoft Approve now also transfers properties from document entries, such as the name or shared category properties.
As a result, additional information – such as that from supplier documents – is also available during review and approval processes. This improves collaboration with external partners and ensures a consistent data foundation throughout the entire document process.
For more information about the key new features in Fabasoft Approve, click here.
