imprezzeo

So Sir James Crosby, erstwhile career banker and chief fox in charge of the UK banking hen house did the decent thing and skewered himself on his own Montblanc. While we have only heard one side of the story, my own brief experience in banking would tend to support his accuser. At LexisNexis, while the US operation already had a huge public record-based risk management arm, I encouraged the European operation to move into the compliance data and solutions market, specifically anti-money laundering / client acceptance. It was only really the prospect of a prison stretch (the Riggs Bank farrago saw to this) that motivated senior management at the banks to demand more assiduous checking of their clients – and the reality is fear can be very powerful purchase driver.

Had the same penalty been in place for lax banking risk management, would we be in this position today? The Basel II Accord and it’s recommendations on capital adequacy were much trumpeted in 2004 as the framework for effective risk management – and little has been heard of it since. So why did Basel fail us so spectacularly? My two cents is that too much was left in the hands of local regulators to a. ‘customise’ their model to suit the local market and b. to apply it according to their own timeframe – a classic fudge.
 

I was invited to an Autonomy lunch / presentation yesterday on how, via their IDOL platform they can optimise SharePoint, especially in large distributed organisations. I can’t help but be impressed by the progress they’ve made and the resilience of Mike Lynch and his team. Deriving meaning from unstructured data might once have been seen as an esoteric search backwater but not now - its the foundation of a £3bn market cap business that looks set to enjoy continued growth. I remember tracking the stock not long after IPO and it took a beating for a while (poor IR programme?, understood but unloved?, a high value / low volume pipeline promising lumpy cash flow?). Look at its performance in the last 5 years v. the FTSE Techmark 100.

Of particular note is the trend-bucking trajectory in the last few months. With Interwoven now in the stable, a balance sheet that would make rivals blush and regulatory and governance imperatives driving strong demand in key markets (e.g. compliance and risk management, reputation management, e-discovery and legal disclosure), their challenges are enviable ones. It may not be the search brand de nos jours and some of that strong potential might already be priced into the stock, but if there’s a better “value stock” right now, I’ve yet to see it.

One of the unique aspects of Imprezzeo Image Search is the way it combines several aspects of visual search technology. There are three layers to the Imprezzeo application. The first layer is a general content search, where image features such as colour, shapes, textures are indexed in Imprezzeo's database and users can match reference images based on feature similarity. Next, Imprezzeo uses face detection to determine the presence of humans faces in an image. Lastly, Imprezzeo adds face recognition to allow a user to search for a specific “persona” using visual matching against Imprezzeo's database of recognised individuals, ie it knows who they are by face and name. Combining these visual search technolgies into an integrated and scalable enterprise search platform is what makes Imprezzeo so appealing to businesses with large image databases.
 

When we initially began talking to people about Imprezzeo, we had to explain these various aspects of visual search, especially the concept of facial recognition. However, with several major vendors beginning to offer various forms of visual search for consumers, the whole concept is becoming much more understood by the general public. Over the past year, Microsoft and Google have introduced rudimentary visual search options into their consumer photo applications. But, it has been Apple, with it's recently introduced iPhoto 09, which includes face recognition, that has caused the most excitement. Not so much because of what it does; its FR technology is just OK, but, in typical Apple fashion, they've developed a very slick user interface that makes it easy for anyone to use face recognition technology. As visual search is used by more and more consumers, we believe it's only a matter of time before it becomes a standard component of popular search engines, content management, ecommerce, and digital asset management systems.