Stay on Top of Emerging Technology Trends
Get updates impacting your industry from our GigaOm Research Community
TempoDB, the Chicago-based time-series database startup that emerged from TechStars in 2012, has changed its name to TempoIQ and has refocused its business around providing a collection of analytics capabilities for sensor environments rather than just collecting their data.
Andrew Cronk, TempoIQ’s co-founder and CEO, says the company started with capturing time-series data for things like sensors and connected devices because it was the hardest problem to solve in those spaces, where companies were historically trying to contort that data into relational databases or other database systems not designed for that use case. And although it was reasonably easy to attract individual developers with the notion of a time-series database service (TempoIQ is a cloud service), demands began to change as the company tried scoring bigger deals such as Silver Spring Networks.
Larger prospects, Cronk said, often had a story that went something like this: “We don’t really have a time-series problem. We have a problem with all these sensors.” They wanted to be able to make sense of the data that sensors are spewing out — that means not just knowing when events happened, but also knowing what they look like visualized, what constitutes normal behavior and what series of events led to any given outcome. In other words, they wanted help monitoring and analyzing all that data.
Hence the change from TempoDB to TempoIQ. Now, Cronk explained, the company enables all sorts of functions on the data it’s storing for customers. Users build applications and they can program the TempoIQ backend to deliver alerts based on a variety of rules, thresholds and event chains. Data is stored at a very granular level for a very long time (TempoIQ ditched Hadoop to build its own highly compressed storage engine) and users can combine it, aggregate and summarize it as they like. Cronk said TempoIQ also enables “virtual sensors,” which are new, logical measurements derived from the data already being collected by existing sensors.
Essentially, TempoIQ is handling the first four steps –capture, storage, monitoring and analytics — in what Cronk calls the Maslow’s hierarchy of data needs. It’s working toward the final two steps of prediction, which implies some understanding of causation (think about what artificial intelligence company Numenta used to boast about), and decision-making, which means suggesting to users what they should do based on the data. “We should and we will [do those things],” Cronk said, “but we don’t.”
However, the new business model isn’t without risk even if it has resulted in some big deals early on. A big one, Cronk acknowledged, is that the new business model means TempoIQ is no longer a developer-focused company. If TempoIQ does handles the shift right, it will be able to pitch CIOs, sell them on a new application to solve a specific problem, close a lot of big deals and live happily ever after. But if sales don’t come easy, or at all, the company risks seeing its developer business gobbled up by peers such as Keen IO (which is working on adding time-series functions to its set of analytics APIs), which could make back-peddling all but impossible.
Cronk is confident in the new approach, though, because experience tells him large companies and government agencies really do want a full-on platform designed to help them analyze all their sensors, and because TempoIQ has taken the necessary steps to deliver that platform. “This is why we exist as a company,” Cronk said, “to push people up the stack.”