Mayor Michael Bloomberg’s proclamation of the NYC Data Week, as Tim O’Reilly noted at the Startup Showcase, opens with the old adage, “In God we trust, all others bring data.” The 14 finalists at the event, one of about 100 public events this week in New York, brought not only data but also top data science skills and enterprising ideas, hoping that competition judges Tim O’Reilly and Fred Wilson will select them as standing “apart from the rest.”
In a very crowded and noisy room, the finalists had about five minutes to convince the judges of the merits of their ventures. It wasn’t easy on the judges either, having to absorb quickly loads of technical and business details from new players in a market that’s just emerging and is not easy to define and segment. Lacking deep technical knowledge—Fred Wilson admitted as much at the awards presentation—they weren’t shy of asking for advice, especially from O’Reilly Media’s Director of Research, Roger Magoulas, as I observed following them around the demos.
Three winners stood “apart from the rest”: O’Reilly picked up Privacy Analytics, Wilson selected Placed, and “the audience” went for Hadapt. Precog got an honorable mention.
Privacy Analytics is protecting privacy in the most demanding sector or “one of the most draconian regimes” as Tim O’Reilly noted, addressing “a very important problem.” Founded in 2007 by Khaled El Emam, who holds the Canada Research Chair in Electronic Health Information at the University of Ottawa, it helps mostly healthcare research institutions to “mask and de-identify data before they disclose it for secondary purposes,” and also provides them with quantitative HIPAA certifications and recommendations. Privacy Analytics claims to “have developed the only commercially available integrated masking and de-identification tool on the market today, and the only tool that incorporates a risk-based approach.” It will be interesting to see if their approach to hiding identities could be applied successfully beyond healthcare.
Wilson said he picked Placed because it seems to be doing something different from what everybody else is doing. Founder David Shim called Placed “location analytics for the real-world,” aiming to “quantify the physical world.” Its mobile phone app has been downloaded by 36,000 users who have opted-in to share their location data in exchange for points they can redeem with various merchants and advertisers that pay Placed for the analysis of the data—location-related insights they can’t get anywhere else either on their own customers’ behavior or on the locations where they operate. Example: If there are Starbucks and Dunkin’ Donuts coffee shops in the same market, 75% of the time Dunkin’ Donuts outranks Starbucks.
Hadapt’s Justin Borgman told the audience how he met his co-founder Daniel Abadi at Yale, where the company got its start as a research project “bringing together the best of both worlds—the performance and functionality of an analytic database with the scalability of Hadoop—into one unified offering." Hadpat bridges the divide between a Map/Reduce programmer and a business analyst who is used to SQL and traditional BI and data visualization tools. Last week, they released Hadapt 2.0 which includes interactive SQL on Hadoop.
In addition to the aforementioned Precog (cloud-based data warehousing and analysis platform), the other finalists were an interesting mix of big data tools and applications: Axemblr (Hadoop in your cloud of choice), Datacratic Inc. (applying machine learning and predictive modeling to real-time consumer behavior), InfoActive, (data visualization of the interactive kind), Mortar (Hadoop in the cloud), Next Big Sound (analytics for the music industry), NGDATA (enterprise data management software), Qunb (one-stop shop for numerical data), Red Carp Studio (analytics products and services), TempoDB, (time-series database service), and WibiData (platform for large-scale data storage, serving and analysis).
Good luck to all! Any bets on where these startups will be a year from now?
