MetaCarta -- A Big Idea

I hopped on my bike last week and rode over to what I believe is the second closest geospatial company to me: MetaCarta. The office is just down the avenue (Massachusetts Avenue) from MIT, where Founder and CTO John Frank continues to study.When I first wrote about MetaCarta in 2002 it was a small company with a big idea.(Joe Francica also wrote about them in May 2004.) Today, the company is larger and the big idea is still valid, and even more in demand.

Recall that MetaCarta's twist on geospatial is its ability to tease out locational references from unstructured data.So, instead of those "clean" addresses in databases used for geocoding, the company's products mine documents for clues in the language that indicate location.Those clues are then geocoded associating documents to locations and vice versa.The first big fans of the technology were in the in the government; in fact, MetaCarta's first funding was from Defense Advanced Research Projects Agency's (DARPA's) Next-Generation Internet initiative and In-Q-Tel, the CIA's funding arm for commercial technology of interest.And, while defense and the military are still users, the company is broadening its customer base, to the commercial and public sectors for new business, as promised from the start.

I sat down with Claudine Bianchi, VP of marketing, and Robert Warren, VP of product management to see how the company is fairing after three years.Bianchi noted about 20-25 customers including those in the public sector (federal, state and local including a pilot in Arizona with the border patrol) and three out of the four top international exploration companies in the oil and gas industry.The experiences with those users, especially the oil and gas industry, led the company to focus on four areas of development: security, stability, federated queries and integration.Warren was quick to note that these are key in any sort of enterprise solution.

MetaCarta has learned that the physical security solution used in the federal government (you are physically prevented from reaching a computer) does not apply in the private sector.There, the name of the game is electronic security.Scalability, in the context of MetaCarta's offerings means speed of "intake" and indexing of the unstructured data, storing the index and data (which may be very large) and ultimately, enabling quickly returned responses to queries.The federated queries allow searching across multiple datastores to retrieve a single answer.Finally, integration means plugging MetaCarta's solution into existing systems.

As the company looks to explore the markets in utilities and other areas, it's found familiar IT challenges: silos of data, often caused by acquisition."It's location," notes Warren, "that can link and cross reference those silos." And, MetaCarta has an "in" here since much of the data is indeed unstructured.He went on to note that the military is seeing the light regarding the use of traditional business intelligence ideas.After running them against the existing structured data, the next step is to run them against unstructured data and then location.

I was a bit surprised to hear that integration was a major development area for MetaCarta.What with all the open systems and standards, plugging one system into another should be simple (and perhaps even invisible).Not so.Just as Bentley and other players in the geospatial arena are developing connectors, so too is MetaCarta.Warren agreed when I suggested that a company like MetaCarta might do well to "stick to its knitting" and hinted that the company is looking to partners to enable these connections, when possible.Still, he notes, "if our solutions can't be integrated with an existing one, there is no growth."

The best news for MetaCarta is that the amount of unstructured data is growing.Despite the use of GPS and even the attention paid to location tagging of Web content, MetaCarta expects to be making sense of unstructured geographic data for many, many years to come.