Thoughtspot Promises Google-like BI Experience
The startup has developed an in-memory database and BI appliance
Wed, February 05, 2014
IDG News Service (Boston Bureau) — Two items on BI (business intelligence) customers' wish lists have long been ease of use and a lower burden on IT. A startup called ThoughtSpot says its new BI appliance can tackle both.
ThoughtSpot has developed an in-memory database for the appliance. This allows information to be held in RAM for faster processing of queries, which users enter in natural language through a Google-like form.
The system relies on the use of structured data sets, in part because ThoughtSpot also provides computations along with retrieving results of queries, said CEO Ajeet Singh.
He demonstrated how ThoughtSpot works using a .CSV file of information on startups from Crunchbase. Singh was able to enter queries that not only brought back subsets of data, such as funding rounds, but also totaled up figures on the fly.
The company is aiming at midsized to large enterprises, which have a variety of structured information sources, such as existing data warehouses and enterprise applications. Customers who buy ThoughtSpot appliances will still have to have their IT departments push data into the system and set up user permissions, but would no longer need to build out reports and dashboards for users given ThoughtSpot's self-service interface, Singh said.
This is the way BI should be, he added. "Imagine if you had to go to a Google engineer to get every [search engine] query built," he said.
While many other vendors are focused on self-service BI, ThoughtSpot's emphasis on a search engine-like experience is a differentiator, according to Singh.
ThoughtSpot's appliance is built with commodity hardware, with configurations starting at four servers. More appliances can be added for scalability. The system uses traditional disk to provide a persistence layer and data backup in the event of a failure.
While a hosted version is on ThoughtSpot's road map, for now it sees demand primarily for on-premises installations, Singh said. Part of the problem with a cloud-based BI service is simply the task of moving terabyte-sized data sets in and out of the system, he added.
Chris Kanaracus covers enterprise software and general technology breaking news for The IDG News Service. Chris' email address is Chris_Kanaracus@idg.com