How it started
We were struck by the number of messages that would go untracked by us regarding various topics. This would later leave us grasping at straws when we wanted to find a vital fact. Even getting to a ballpark for that fact was difficult at times. What we needed was someone like a butler, who could keep us informed of what we needed to know as well as find appropriate facts at the right time.
What can Ariv do for you?
Ariv.ai helps you stay organized and find information that you need from conversation channels and messages. Just ask don't search. You can also tell Ariv what interests you and let Ariv sift through public streams of messages and ensure that you don't miss the ones you are interested in. Ariv also recognises knowledge contributed and knowledge curated and rewards users with karma points.
Under the hood
Our platform iEngage.ai takes in conversations and creates a Knowledge Graph about your enterprise. Ariv.ai, our app, provides an integration layer between iEngage.ai and Microsoft Teams that passes on conversations within your teams to iEngage.ai. The knowledge captured is then surfaced in Teams by Ariv.ai in various ways through its features.
Challenges we ran into
One major issue was getting permissions to listen to channel messages specific only to that team as getting permissions for the whole tenant would be a lot of redtape to cut through all at once. This was made possible by the Resource Specific Consent which became generally available recently.
Other miscellaneous thoughts
Knowledge creation from conversations is far from solved. However, if stakeholders iteratively instruct our system on the quirks of their enterprise and its domain, we'll get closer to a state of man & machine augmenting each other. We're hoping to amplify these feedback cycles further with your help. We're delighted to see that there's a wave of interest to know how knowledge and systems that deal with knowledge can help productivity & engagement. We've been pleasantly surprised by Microsoft's Bot Framework SDK which is open source and adaptable to different environments. We're also seeing the Microsoft Graph API being added to and developing quickly which has helped our use case.
What's next for Ariv.ai
We are currently building out more boards which will help visualise, analyse and formalise your knowledge. More on that later!