Original URL: http://www.theregister.co.uk/2010/09/24/mimecast_babk/

Mimecast digs its way into the information bank

Outside the vault box thinking

By Chris Mellor

Posted in Storage, 24th September 2010 14:25 GMT

Analysis Mimecast is recasting itself as an information banker and looking for new types of data it could store and new ways to mine the banked information.

The company is a UK-based concern offering e-mail as a service and working to add the storage of unstructured files and SharePoint information to what is an e-mail capture, indexing, search and archive platform offering.

It has a chief scientist, Nathaniel Borenstein, who was hired three months ago and was previously an IBM distinguished engineer. He joined the company to achieve things he couldn't do at IBM. There, the corporate bureaucracy and excessive company size meant that it was virtually impossible co-ordinate all the relevant company resources on topics such as SPAM control. During 2006 he had a meeting set up between IBM and Symantec to discuss an anti-SPAM alliance and found out IBM had bought Symantec's then biggest competitor Internet Security Systems on the day of the meeting, rendering it pointless and embarrassing.

Mimecast, he says, is a company that is fast-growing and past the early start-up phase with its 150 or so employees, yet small enough for left hands to know what right hands are doing.

He owns, or can look at, all the technology not owned by co-founder and Chief Technology Officer Neil Murray. We might say he is licensed to think outside the existing Mimecast box.

It's a truism that information is today's business currency. Borenstein considers Mimecast is like a bank that takes in customer information deposits. He is looking at capturing and storing new types of information on the one hand, and new ways of mining it on the other.

"Archiving e-mail is fixing a big enough pain for customers. Suppose, you could do ... telephony archiving in the same way as email. ... You could run phone messages through voice recognition and transcription processing and get rough voice speech records ... and have them searchable."

Storing voice-over-IP (VOIP) calls would be a little simpler as they are digital when they enter the system. He says: "HPPA data requirements don't generally apply to phone calls because phone record archiving was not thought to be possible when HPPA was conceived."

Borenstein reckons telephony recording might be a good idea for lawyers and finance trading companies: "Lawyers could choose it in advance of regulation." In the finance world regulators might find it very desirable to have a record of all traders' business calls when investigating market irregularities.

He is wondering whether this would be a service worth selling, and thinks it would be a Pandora's box-type thing. Once it was out in the wild you couldn't put it back: "Mobile operators or telcos could make it available and you won't have any say in it."

The water cooler guy

Another thing you could do is mine stored data for new patterns of information, such as discovering the social network connections between employees: "You could use email data to generate social networking maps."

Some people can serve as a communications hub, with lots of emails coming to them and then going out to other people: "Those people can get terrible peer reviews because they don't accomplish anything else. ... But they are valuable."

We might say that in a distributed and connected world they are the virtual water cooler. Other people have conversations by and through them: "A social networking map could reveal this ... could reveal people who are key connectors."

This technique could reveal other patterns, like the SPAM problem at IBM where there were lots of people "working on it and not talking to each other. Such functionality could bring these efforts out and make the links."

Bottleneck relief

Borenstein reckons Mimecast could have a development bottleneck with too few developers for all the developments that could be desirable. One way to fix this is by developing an Application Programming Interface (API).

"It's not easy to free up resources to implement good ideas and not easy to hire external people." So he is working on the idea of a generalised API, "Which could enable apps to be developed to use data in Mimecast and the app developers could be both in and outside Mimecast ... as long as they had the right permissions."

This seems very Lotus-like with its idea of an expanded virtual team of developers inside an Active Directory or similar policed corrall.

Mimecast recently ranked second in the UK's Sunday Times Tech Track 100 list of the fastest-growing privately-owned technology companies in the UK. Hiring Borenstein and Borenstein's ideas show the directions the company is looking at and indicates products or services that could hit the streets a few years hence.

Knowing that every word you said on the telephone at work would be recorded could have a real sobering effect and make off-site water cooler meetings more desirable, but then Borenstein's social media mapping could find their digital footprints too. Mimecast could become a digital tracker in the cyberspace jungle. ®