Turning the Internet into a corporate database

The title of this piece is not based on my words but those of another analyst describing Bitvore. So, what does this (the company and product name are the same) do? I can best describe this through the example of a current prospect of the company’s. This potential user grows salad and vegetable products and it needs to decide what to plant nine months before its produce is ready to be harvested, packaged and sent to market. Not unreasonably it wants to know what are going to be the ‘in’ salad and vegetable products next year: will it be kohlrabi and radishes or cabbages and tomatoes? In order to try to predict this it is talking to Bitvore about monitoring the menus published by restaurants online as well as recipes that are published online. Then you can start to trend that information so that you can see what is becoming more popular and also identify who are trendsetters in this market and what they are getting excited about.

As it happens most of Bitvore’s current customers are with local US municipalities, monitoring the Internet for information relevant to local municipality bonds. As this is a somewhat esoteric market that will not be understood by most readers I thought that the above example gives a better understanding of what Bitvore can do. But I probably need to go into a bit more technical detail.

Bitvore is a cloud-based service where each customer has a dedicated private cluster that uses HBase (which is also used to store all relevant metadata for governance purposes). You can also have remote Bitvore instances, either for back-up purposes or for privacy reasons (so that information does not leave a particular constituency). The product itself uses natural language processing and has a tagging capability that looks like Google Adwords. What you get presented to you will obviously depend on your requirements but in the case of municipalities, for instance, they will get a prioritised list (typically top ten) of the latest articles or news about specific bonds. In other words it works like Google but with context, so that what you get is stuff that you know that you are interested in.

If you want to do sentiment analysis then there is a plug-in for Alchemy but, more generally, you can monitor tweets—not for this purpose but to identify relevant URLs that are mentioned and then Bitvore will access these URLs to see if the results should be returned to you as one of your top ten (it could be 5, it could be 20: you decide). Other collectors are available for RSS, websites, various social media outlets, SMTP/iMAP and Bing. These are extensible plug-ins and new ones can easily be added.

There are also various “Analyzers” included within the product, for example there is one that de-duplicates information so that you don’t get the same reference twice and another that analyses geographical information (for example, where a tweet came from).

The company is actively seeking systems integrators and partners with domain specific knowledge to build relevant applications on top of Bitvore. For example, one partner has built a solution for motor companies to manage their dealer networks, which effectively provides an extended 360^o view (which, as an aside, is a growing market) across its dealerships.

I have to say that this is a nice product that looks easy to use. I haven’t seen or heard about anything else quite like it and it appears to fulfil a clear market need. As an alternative to “turning the Internet into a corporate database” another analogy (credited to one of Bitvore’s customers) is that information on the Internet is like a river of information flowing past and they want to pull fresh data from the river and turn it into an asset. However you want to describe it, Bitvore is definitely worth a look.