When our demo prospects ask, "where's the A.I.?" in Talla, what they really mean is, "where is the magic that I didn't think regular software could do?" In truth, A.I. is all over Talla today, but we have plans to create a "virtual managing editor" in Talla that will feel pretty magical -- and it's coming soon.
The current version of Talla's knowledge base A.I. is largely focused on problems of information discovery, which is to say it understands content structure and search terms in a more sophisticated way so that it makes finding the right answers easier. This is perhaps the most important role A.I. can play in your knowledge base, but it suffers from one glaring problem -- Talla's A.I. can't point you to the right answer in your knowledge base if the right answer isn't actually in your knowledge base.
More simply: search can't solve bad content. And if your employees can't find the content they need in your knowledge base, the knowledge base will be worse than useless -- it will be unused.
That's where the virtual managing editor will come in, as it will help you determine what information is missing or out of date in your knowledge base, and then take steps to help you fill in those gaps. The managing editor A.I. will point out where search is destined to fail, and help you correct for that lack of content.
The first job of a managing editor is to organize your existing content into its most useful form, such that casual human readers or search algorithms can most efficiently browse your content. (The technical term is information curation.) The managing editor A.I. will perform a topic modelling exercise that will suggest an optimal category and tag structure for your knowledge base content, rather than simply relying on the taxonomy that's already in place.
The managing editor A.I. will also be semantically aware of all your content, and be able perform Named Entity Recognition to properly tag persons, locations, and other significant nouns in your content, which will limit or eliminate the need for a human to tag anything.
The virtual managing editor A.I. will automatically observe which of these significant persons, locations, and nouns are undefined in your text, and suggest what new content needs to be generated (as in, it will help with information creation). For example, if the job role "travel coordinator" keeps popping up, but nowhere in your knowledge base or attached organizational chart is the travel coordinator identified, the editor A.I. would flag that as needed content.
Moreover, once the travel coordinator is identified, and content is associated with that job role, the managing editor A.I. can assign contents task to the travel coordinator. If travel policies fall out of date, or the org chart is updated with new person in the travel coordinator job role, several content items can be surfaced for updating and improvement.
Extend this process to every job role and every piece of content, and your knowledge base will never "grow stale" because the virtual managing editor won't let it.
The Talla A.I. of today will ensure the best possible information discovery for your knowledge base. The Talla A.I. of the near future will ensure that information curation and creation are optimized, too. And when a software agent can act like a professional managing editor for your knowledge base, that's a form of A.I. magic that will be obvious to everyone.
Want to get an early look at what A.I. can do for your knowledge base -- both today and tomorrow? Contact Talla today!