The Offical FatBrain Blog

Announcements, discussions, and more from FatBrain.

Google's GCP Next Signals Hyperscale AI for the Enterprise

Modern Enterprise Architecture means Software that Learns Business Outcomes.

 

March 26, 2017

 

The FatBrain team participated in Google's GCP Next global annual innovation event in San Francisco earlier this month (link). In case you missed it, here is the Lead Theme of the event, curated by FatBrain's Cognitive Themes engine: "Modern innovation architecture means software that learns business outcomes and improves them over time." Videos of all sessions from Google Next 2017 can be viewed here. Several additional items are also worthy of notice.

 

Google's AI business forms a critical part of its strategic cloud business growth, led by Diane Greene, co-founder and CEO-emerita of Vmware. This highlights Google's long-game corporate commitment to enterprise, poignantly and pointedly reaffirmed by its executive chair Eric Schmidt, in the keynote address (link).

 

Google has made it a corporate priority to differentiate its cloud offering by infusing the AI-driven Design Thinking across its platform. This includes not only opening up its OSS patronage projects for deep learning frameworks like TensorFlow, but also making the underlying GPU-enabled compute layer available for just in time and server-less consumption (link).

 

Google is using its price and scale advantage for G-Suite (e.g., gmail, drive, docs) to penetrate enterprise accounts in direct competition with MSFT Outlook 365. When combined with machine learning advantages (such as FatBrain's Cognitive Themes), this could form the early AI-based wins for corporate clients looking to modernize their communication infrastructure (link).

 

Together these offerings play a key role in Google's strategic drive to become an AI-first company where algorithms enable the automation of the Scientific Process in every facet of their business. To that end, Google showcased multiple examples replacing complex, expensive rule-based systems with adaptive, self-learning AI software. In each case, the principled AI systems significantly outperform the hand-crafted and hard-to-maintain traditional rule-based approaches, very much inline with the FatBrain's Cognitive Themes results (link).