Here’s an example:
With the “Find the Expert” insight application, end-users perform search queries to find subject matter experts (SMEs) on a particular topic.
In this insight application, users can perform natural language queries. Let’s use this query: “Who is an expert on diabetes?”
This search gets dissected using NLP to extract the core meaning behind the query, considering every word and entity (who, expert, diabetes). In return, the insight app knows you are looking for a person, in this case, contact information for an SME on diabetes. NLP’s understanding of the query gives the user a name, phone number, email address, and even more in a matter of seconds – the fastest and most relevant search results.
Users are also able to submit their queries using speech processing. This voice-to-text plugin allows one to use voice recognition and enter their questions hands-free via a microphone. After this voice-to-text input, the query is treated the same way as a textual query and is subject to the same NLP plugins mentioned above.
There is even more!
On top of this, end-users may also query with multimedia input. For example, an image can be uploaded directly in the search integration then sent to Mindbreeze InSpire as a query and processed. Once this multimedia query has been processed and understood, the data extracted can return highly relevant search results related to the multimedia – this includes related images, visually similar images, or documents that mention objects recognized directly within the image.
In addition to multimedia, complex data is accepted as input for queries, such as CSV and spreadsheet data files. These data sheets can either be processed wholly or iterated over to produce a range of search results. One concrete example where this is utilized frequently is with our Bid Automation feature. This feature processes a submitted Request for Proposal (RFP) or Request for Information (RFI) excel spreadsheet containing questions. It then returns relevant knowledge base articles for each question and delivers the answered document fully.
This automated process is also powered and made possible via our NLP pipeline to identify the questions within the submitted document and find the relevant answers from the indexed knowledge base articles.
Integration into Various Applications
When Mindbreeze InSpire is integrated directly into an application, such as Tableau, Power BI, Cognos, and more, the available data is hooked instantly into the Mindbreeze InSpire integration. This data, taken directly from the application, can also drive search queries sent to Mindbreeze. For example, end-users can directly interact with their data visualizations in Tableau and immediately see this reflected in the search results of that data within Mindbreeze InSpire, allowing for seamless and immediate cross-platform data presentation.
Mindbreeze InSpire can also be used in workflow automation and streamlining. Users can send whole documents as queries to Mindbreeze InSpire via an embedded insight application within their workflow process – our solution will further process and add additional insights to the document. One example would be taking the uploaded document and utilizing the Mindbreeze semantic pipeline to recognize and extract a contract value. Once this has been identified, the document can be sent further down the workflow for processing by the relevant people or departments based on the contract value information.
Similarly, in case deflection use cases, Case Deflection Claim Forms can be submitted to Mindbreeze InSpire as a direct query. This claim form is processed by our application and the NLP and semantic pipelines to understand the problems presented within the document. With this information at hand, the most relevant knowledge articles and FAQ answers can be delivered to the end-user.