Herb Cunitz, the new president of HortonWorks, is an Open Source veteran. A sales and marketing team builder and leader, he was head of worldwide field operations for SpringSource. When VMware acquired the development platform startup, he moved to VMware, where he ran global field operations for the vFabric Cloud Application Platform division. SpringSource, he said, “at the time was a chance to have a transformational impact on open source development…. I see the same opportunity for transformational change in data with Big Data right now. Open Source will help change the face of data. Hortonworks has the opportunity to do that.” That, he said in a Cube interview from the Strata + Hadoop World 2012 conference in late October, was what motivated him to leave a successful and challenging position with a very fast growing company to join a startup at the sharp end of Big Data infrastructure development and implementation. Part of the excitement, he told Wikibon’s David Vellante and SiliconAngle’s John Furrier, is that this is being driven by Open Source, and Hortonworks and its principals are all major contributors to the Open Source community. Hortonworks invests its profits back into that community in the form of advances in code surrounding Hadoop, all of which are donated back to the Open Source community. SpringSource, he said, followed a core model, building proprietary extensions on an Open Source core, in its case a development platform based on Tom Cat. Since then, however, the market has changed. “In some ways we are in Open Source 2.0.” Today that market wants all the code to be part of Open Source to avoid ending up going down technology forks. So instead of selling a semi-proprietary platform, Hortonworks makes its money providing service to Hadoop users. “Customers want to know that a trusted supplier will support them with standard commercial terms – one hour response times, etc. – if something goes wrong…. If the community is large enough and adoption is broad enough, there is plenty of money there.” Partnerships Versus Platforms The other core element of the Hortonworks business plan is partnership with existing players. Rather than trying to push its own Big Data platform as the “answer for everything”, it works with established data analytics providers to incorporate Hadoop into their products. Recently Hortonworks has announced two major partnerships, with Teradata and Microsoft. This again is in part a strategy for keeping all the code Open Source. If these vendors had chosen to develop their own Big Data extensions to their existing products, they would inevitably have created proprietary forks in the technology. It is also the best way to introduce Big Data and Hadoop into company operations. “For example, Teradata is taking our distribution and embedding it into its product as they take that to market. So I’m a Teradata user. I know how to do queries through Teradata, and now I can use the same tools to access Hadoop data. Or I’m an administrator managing the console. I know how to manage everything around Teradata. Do I really want another pane of glass to look at to manage my environment, or do I want to go to the same place I’m used to but now also manage the Hadoop environment?” By working with Teradata and other partners, Hortonworks makes it as easy as possible for their clients to introduce Big Data analysis into their environments without the expense and problems of training their employees in a new technology. The Microsoft announcement is particularly exciting, he said. “The Number one BI tool on the planet today is Excel. If I really want to bring Hadoop to the masses, I I can say an Excel user can get access to his normal SQL data and bring in Hadoop data and through the same UI can manage, organize, and enrich that information with Big Data and visually present it back in a way to get insights, all through Excel… that’s the power of the Microsoft announcement.” All of this, he says, ties back to the philosophy of “helping to make Hadoop easier to install, manage, use, develop against, and consume.