4 Comments

Summary:

Zoomdata has a plan for business intelligence that involves tacking the difficult problem of streaming data, and doing so with a mobile-device-first mindset. The result is pretty and compelling in theory, but it’s technologically challenging and will face tough competition from new and old vendors alike.

StartupInfluencers_Full_101812

Take streaming data, then sprinkle in some Hadoop, an array of visualizations and a user experience designed for touch screens, and you have Zoomdata. The Reston, Va.-based company launched on Tuesday with $1.1 million in seed funding and a mission to prod business intelligence into the mobile-first world.

Zoomdata Founder and CEO Justin Langseth started Zoomdata on the premise of building a company that envisioned BI free from decades of legacy baggage. In 2012, that means abandoning the desktop and designing for tablets, and taking advantage of the nearly unlimited computing power available in the cloud and even on our mobile devices. It also means designing a user experience so intuitive that users know how it works without ever really having to learn.

Just like someone can open up Google Earth and know they’re seeing the planet Earth, Langseth said Zoomdata users should open the app and say, “Oh, that’s my business.” And then they should be able to easily zoom in right where they want to go, using only their fingers. In a few swipes and pinches, Langseth said, users are soon uttering the business equivalent of “that’ my house, that’s my car, that’s my tree.”

It’s a pretty heady concept for a guy like Langseth who has been entrenched in the space for years, first at MicroStrategy in the 1990s and most recently doing a text-analysis startup, but he appears to have pulled it off thanks to the array of powerful components now floating around the web for free. Zoomdata is able to take data from anywhere — web apps, enterprise systems, Hadoop, email, you name it — and process it as it hits the system using an open source stream-processing engine called Storm. Once it’s processed, Zoomdata applies intelligence to figure out the best way to display that data visually and puts the result on the screen.

Data is like paint on a palette

That, Langseth said, is where the magic really comes in. “We’ve been thinking of data as kind of like paint,” he explained. The app takes many of the concepts from the D3.js project for creating HTML documents using data, but then makes them interactive and “lights them up with real-time data.”

The human interaction becomes a combination of watching a movie and finger painting. Combining data sources and sets by swiping your fingers is akin to blending colors from a palette. The interface comes with set of buttons for pausing, rewinding and fast-forwarding the visualization, too — because it’s a real-time engine, the data keeps coming and the visualization keeps changing until someone temporarily stops the flow.

Under the hood, of course, Zoomdata is a lot more complex than meets the eye. It’s all about scale, speed and huge amounts of data, Langseth explained.  The backend does all the work and only streams the data required at any given time, so as to save the processing load on the user’s device. If a user presses pause or rewinds, the system keeps processing new data while also letting the user interact with the older data unaffected. Zoomdata also supports historical data sitting inside databases and other data stores so that users can compare their real-time information against the past.

The future: Bigger screens, smarter visualization and stiff competition

As if all this doesn’t sound futuristic enough, Langseth’s plans to take the technology further. “There’s a whole bunch more intelligence we can add to the system,” he explained, referencing his plans to incorporate machine-learning algorithms that will make the system even better at choosing how to visualize the raw, often schemaless data it’s receiving.

He also likes the idea of big touch screens, like CNN Big Board big. Sometimes when he has his iPad display showing on this 50-inch office television to play music, people come in and just assume they can start interacting with it like a touch screen. “Not just being able to see it, but to touch it, really excites people,” Langseth said.

Of course, as with all companies trying to carve out their space in the lucrative BI market — including the newly cash-rich Platfora — Zoomdata will have to prove itself a worthwhile alternative to big, expensive legacy technologies. Langseth thinks the real-time, mobile nature of his company’s product will at least make it a nice complement to existing desktop-based BI tools for historical data. And like pretty much everything powered by the cloud and rendered on a mobile device, its simplicity might appeal to a lot of users who don’t need the price or complexity that comes along with much legacy software.

“Some people need 10,000 features,” he said, “but most people need 5 features.”

We’ll see if Langseth is right soon enough. Zoomdata is currently in private beta after development began in March, and the company hopes to keep refining the user experience and open it up for broader consumption next year.

  1. If you want a data visualization platform on iPad check out pixxa.com’s Perspective native app, they are the future as Zoomdata is just an HTML port, still thinking the old BI dataviz model.

    Very impressive stuff. more for presenting/sharing data that is Oh so easy!

    Share
  2. “But most people need 5 features.” Very interesting ideas. When moving to mobile, UX matters so much more and getting the right features matters far more than all features. This is a trend I expect to see continuing across any platform that integrates with large data and, even more so, ubiquitous data.

    What’s also interesting is the amount of heavy lifting being done on the back end so that the mobile device can keep up with the information stream. This is an integration story more than anything, as moving the key data fast requires a host of infrastructure pieces that have little do with Hadoop.

    BI is dead. Long live UX-focused BI, whether mobile-first or not.

    Share
    1. Chris – spot on comment. That no integration vendor has yet claimed the “mobile first integration suite” leaves me a little surprised. The thing is, mobile-first is important when it comes to an approach, so it does matter if things are optimized for mobile or not (not sure if your last sentence meant that mobile doesn’t matter).

      Share
  3. Joshua Danehower Monday, November 19, 2012

    Anyone know how you would use a Machine Learning Algorithm to decide what the best type of visualization is? That’s very interesting to me.

    Share

Comments have been disabled for this post