Google Cloud exec talks courting enterprises, competing with Amazon and Microsoft

Google kicks off NEXT user conference this week in San Francisco

1 2 Page 2
Page 2 of 2

That very much describes me in my last company: You have a certain set of skills and priorities that are almost always dominated by keeping the lights on and carving out the resources to do net new things is hard. So what we advise companies is to prioritize the use cases. You can try to do things in a big bang way if you’re really committed to it, but more often than not you really need to map out a journey over time. It’s important to have a path to go to the cloud – it doesn't have to be decades, it can even be years – but that will allow you to demonstrate proof points along the way.

Another important piece is to really automate and leverage tools that allow you to remove the operational burden of operating your IT environment. A lot of customers use Google App Engine because they’re saying, ‘we don’t have to deal with all the operational pieces that make this run, we just want it to work.’ One of the customers we’ll have on stage at NEXT will be talking about how they basically moved all their apps into App Engine and that freed up a huge number of technical resources that they’re now redeploying on machine learning, data analytics and refactoring apps that are harder to move. Having a very thoughtful migration journey that takes into account the skills and needs of the app is really important.

What about machine learning, specifically? How can customers begin to implement this technology, which I get the sense that a lot of people believe is pretty futuristic?

We think it’s very much a today technology and it’s getting better and better over time. We have two different approaches to this market: There’s a lot of very sophisticated companies that have data scientists who have the beginnings of a machine learning team, so we provide Cloud ML, our managed environment for machine learning to really make that an easier process. It automates processes that don’t provide a lot of value and provides compute processing to power them. If you want to do something bespoke, we think Cloud ML is a terrific platform for doing that.

On the other side, we’ve got what are essentially pre-trained models that Google has built; an example would be our translation models, image recognition models and natural language processing tools. Many of our Cloud ML services run on custom Tensor Processing Unit chips. We make these services available to any company on an API-basis. This means that if you have a large image library within your company that you want to classify and make searchable, you can use our off the shelf machine learning to basically accomplish that task, you don’t need machine learning scientists inside the company. We’ve done all the hard work and make it available for you to use. We’re trying to play on both sides because customers are in many areas across that spectrum.

We’ve gotten a lot of demand for something in-between, so we’ve created something in Mountain View, and soon in other places, named the Advanced Solutions Lab, which is a co-working space where companies can send their data, analytics and business teams to sit side by side with our machine learning experts and co-develop the models that the company is interested in. We’re trying to build a training capability role and provide engineer-to-engineer connection for people who have compelling ideas in machine learning, but they just don’t know how to make it real.

Join the Network World communities on Facebook and LinkedIn to comment on topics that are top of mind.
1 2 Page 2
Page 2 of 2
Take IDG’s 2020 IT Salary Survey: You’ll provide important data and have a chance to win $500.