Adam Kocoloski
- CTO, Cloud Data Services
- IBM Analytics
- BS Physics and Philosophy, University of Dayton
- PhD Physics, MIT
Adam traded being an entrepreneur for an intrapraneur when IBM acquired his company, Cloudant. Now he’s building IBM’s data and analytics platform in the cloud, and bringing a renewed embrace of open source technologies and development practices.
In his own words
… on going from acquisition to Fellow in less than two years
I founded Cloudant with two of my fellow MIT lab mates in 2009, as a database-as-a-service for developers to build apps. So it was a fast five years when IBM came into the picture. But it was a good thing. It was great to be at a company with a research division. At a startup, your long-term view is next year’s plan. Joining IBM, we have the opportunity to invest in things that take longer and are potentially more disruptive to the industry. That’s something I’m getting more involved in.
And I think the work we did at Cloudant has had a big impact on helping transform IBM into a cloud company. In addition to being a 24x7 cloud service, our team has actively contributed to making IBM an open source community leader.
As for being a Fellow, you do your best to drive the company forward and at some level that becomes noticed. For me, I happened to be at the right place at the right time — at the intersection of open source, cloud services and systems of engagement when it was cutting across many of IBM’s strategic priorities. The best path to Fellow is the one you don’t actively pursue!
… on IP versus open source
IBM’s patent portfolio is a critical part of R&D but there are other ways to accrue intellectual properties, such as being a leader in open source. You’re seeing that influence now in how IBM evaluates candidates for senior technical positions. In many cases, a strong track record in open source is just as valuable to the company as a strong patent portfolio.
… on cognitive cloud solutions
My team provides the foundation for the development of many of these cognitive solutions that require expertise in things like neural networks design and sophisticated machine learning processes. For example, we’re finding that Apache Spark is an increasingly popular and powerful tool for doing machine learning on large data sets. So, IBM has embraced Spark as the foundation of its next generation platform, and we’ve already delivered a fully managed service for Apache Spark on IBM Cloud.
… on where to start coding
It has never been easier to participate in open communities. If you look at GitHub today, more than two million repositories have seen activity in the last month alone. It’s unbelievable. Bits of code are now more discoverable, more consumable, and it’s easier to participate. You can find a project that interests you, contribute code, and submit it to be included.
I publish my own code on GitHub, and all of the key open source projects on which we depend are available there. A senior member of our team spearheaded the work in the Apache Software Foundation to start mirroring all of the code repositories to GitHub. It’s been a great way to encourage new contributors to join our projects.
IBM also has an internal GitHub install courtesy of our CIO’s Project Whitewater. We’re directing all new development to be hosted there. At a large company like IBM it’s often the case that open source development practices can be just as effectively internally. Tim O’Reilly coined this “innersourcing.”