Vertica is a fabled company in the analytics space. It was a pioneer in the column store data warehouse world and one of many companies in which database pioneer Michael Stonebreaker had a hand in founding (Tamr and VoltDB are just two more recent examples).
In Vertica's salad days, and into the HP acquisition, Chris Lynch manned the helm as CEO. He was also an investor in sqrrl, DataRobot, and Hadapt, among other companies. As a principal at incubators Hack/Reduce and Hack Secure and having raised two funds at Accomplice VC, Lynch looms large over the Boston Big Data scene.
Under new management
Clear across the country in San Mateo, CA, lies AtScale, a company founded by Dave Mariani who, after stints at Yahoo and Klout, was frustrated both by the user-unfriendliness of Hadoop and the scaling limitations of OLAP (Online Analytical Processing) systems like SQL Server Analysis Services. In response, Mariani built a company and a product dedicated to providing a scaleable OLAP system that ran on Hadoop.
Also read: Is this the age of Big OLAP?
Starting today, Chris Lynch is AtScale's CEO. Mariani will serve as Lynch's "technical partner," much the way Stonebreaker did at Vertica, according to what Lynch told me in a telephone interview last week. Lynch will be staying Boston Strong, which is to say he will not be relocating to the Bay Area. While Lynch told me he'll be traveling so much that his home base shouldn't matter much, his track record and his evident Massachusetts accent made me think he could be more effective in his home element.
A gradual approach, and not-only Hadoop
Lynch is excited about AtScale, and for the perhaps counter-intuitive reason that its product is more evolutionary than disruptive. Lynch says AtScale can first run over on-prem data warehouses and then on their cloud counterparts, all while maintaining a consistent interface and access to support popular BI tools. Lynch feels that such a staged approach, which helps avoid a "rip and replace" strategy, is a key catalyst to helping Enterprise customers start and complete their journey to the cloud.
The careful reader will note that I first described AtScale as running on Hadoop, but that Lynch discusses its ability to run on data warehouse platforms. In fact, AtScale is now capable of running on both. And Lynch specifically told me that he aims to make sure that the percentage of AtScale customers running the product on Hadoop will become proportionate to Hadoop's share of the analytics market, relative to other platforms.
Also read: AtScale 6.0 and Kinetica 6.1 announced; SAP gets NVIDIA GPU religion
Changes and silo-busting
That would be a big change for AtScale but Lynch clearly feels strongly about it. Can he make it so, though? Considering Vertica created an in-memory OLAP platform on a relational data warehouse, bringing analytics customers back into the SQL fold, from where it believed they never really wanted to leave, the same could come true here. We'll have to wait and see.
Regardless, though, this move will help break down the divide between the data warehousing and Big Data worlds, in terms of geography, culture and Enterprise vs. startup customer platform decisions. It will also bring maturity and deep industry Enterprise expertise to a startup that wants to rise to the proverbial next level.