Oct 7, 2012

Skype Meets Siri: Is MindMeld the Future of Collaboration?

Shouldn’t enterprise collaboration be more than just adapting consumer technologies for work (‘Facebook for the enterprise’)? Most of the 'Social Enterprise' tools are for asynchronous text-based communication, while the most important type of (remote) collaboration in enterprises continues to be real-time by phone/ video conferences. Hence it was fascinating to see an advanced voice recognition & search solution from the San Francisco start-up Expect Labs at the TechCrunch Disrupt conference a few weeks ago. They were one of the finalists and according to Robert Scoble, the hottest iPad app at Disrupt!


'Skype Meets Siri': is how the company referred to themselves, though I personally felt it was more like ‘Project Glass for Voice’! It is a voice-video calling app (built by Tim Tuttle - MIT Ph.D. in Artificial Intelligence - & team) that understands conversations in real-time and pulls up relevant information - pictures, videos, articles - in a ‘Flipboard-like’ interface. The magic is in using AI algorithms to decipher and predict implicitly mentioned concepts (unlike an explicit question to Siri) and then extract information from the web, social media, and in the future, enterprise systems such as CRM. Results can be effortlessly shared in real-time with participants and the snippets become the meeting notes & follow-up items.

Enterprise Tool: MindMeld today is a consumer app, but I can’t imagine that every casual conversation with friends needs such a sophisticated tool. I think it’s far more suited for an enterprise context, since a big difference in how consumers interact and enterprise employees collaborate is that most business gets done verbally, while consumers are happy to spend a lot of time texting or typing. Real-time verbal discussions (face-to-face or over phone/ video) will continue to be the most important form of collaboration at work.  

Why Enterprise? Sales people practically spend their lives on the phone and only send terse emails/ IM/ chat messages. Adoption of social tools at work has been slow due to the extra effort & information overload. In the busy work world, executives hardly read long emails and often don't type much text (on email or 'Enterprise 2.0' tools) due to the risk of misinterpretation or liability issues. (This might also explain why executive participation in internal social tools is still limited). Document & text-based collaboration tools will always be valuable, but it could be complemented by innovations such as voice recognition to make life easier for employees.

Examples:
Imagine a sales call, where you are presenting your products to a prospect. He then starts talking about competitor products that you might not even be aware of. MindMeld could pull up information in real-time (e.g., spec, product comparison or an internal blog post on your CRM knowledge base), without any explicit and frantic search. You can confidently show (selected) information to your prospect and continue your sales pitch. It's like having a virtual faithful assistant (like your best Inside Sales Rep) constantly helping you behind the scenes supplying vital information that makes you look good in front of the customer - even answering questions that you didn't even know you had! 

Future - Here, but Not Evenly Distributed: There are many enterprise scenarios for real-time collaboration such as customer support/ call center applications or learning. I think MindMeld is a feature and not a standalone product, but that feature could make every conference, where people have limited prior context and time, more productive. Philosophically, MindMeld might also serve the noble goal of connecting people based on common ground and empathetic listening (though outsourced to a machine!)

The best quote I heard from the Disrupt conference was “AI is the new UI”. The new age of artificial intelligence might be taking evolution to the ‘singularity world’ pretty soon :-)

A 2-min product video tour of MindMeld:




P.S. I just saw that MindMeld got $2.4M funding from Google Ventures, Greylock Partners etc. Congratulations to the team and all the best!