Facebook recently acquired AI startup Ozlo to add more capabilities to the Messenger product. Ozlo technology is likely to also support the intelligent assistant Facebook M. That product has reportedly struggled to transform its human assisted AI into an independent virtual assistant. Ozlo’s website content was completely deleted and only a message titled “We’re Joining Messenger at Facebook,” remains.
According to Crunchbase, Ozlo had previously raised $14 million in one round of financing from Greylock Partners, AME Cloud Ventures and Yahoo! founder Jerry Yang who invested both personal funds and as part of AME. Ozlo was founded by Charles Jolly, a former Facebook employee. TechCrunch’s John Mannes reported yesterday that Jolly will not be returning to Facebook as part of the acquisition.
Facebook Intensifies it’s AI Efforts
Voicebot reported last week that Facebook was planning a smart home intelligent assistant device, much like Amazon’s Echo Show, for launch in 2018. Facebook Messenger has over 100,000 chatbots and a similar number of developers, according to comments by Facebook’s David Markus and as reported by The Register. These chatbots and Facebook M can certainly benefit from Ozlo’s text analysis and knowledge graph technologies. An example from an earlier version of Ozlo’s website shows how the system can analyze text to extract useful information and return optimal answers for users. The example below finds the lowest price for a video stream of the BBC show Sherlock.
This is not about a search engine that offers you 10 or more links to sift through. This is SEO that returns one or maybe two results. These could be the optimal responses, but it definitely focuses a user’s attention on specific links or in this case, merchants. You can see how this would save a lot of time for all of those human curators at Facebook M. It may even make them unnecessary.
Ozlo Can Be Applied to Voice as Well
While Ozlo is a text-base solution today, it is not hard to see how it can be applied to a voice assistant. Facebook only needs to implement a competent speech-to-text (STT) engine that can offer Ozlo raw text to mine for the optimal responses to user queries. Facebook may be trailing Amazon, Google and the others today in some aspects of voice assistant technology, but acquisitions like Ozlo can help it catch up quickly.