In one of those Autonomy announcements that seemingly appear out of nowhere, the company has declared its intention to “transform” the relational database market by applying its text analysis technology to content stored within database. The tool is called IDOL Structured Probabilistic Engine (SPE), as it uses the same Bayesian-based probabilistic inferencing technology that IDOL uses on unstructured information.

The quote from CEO Mike Lynch grandly proclaims this to be Autonomy’s “second fundamental technology” – IDOL itself being the first. That’s quite a claim and we’re endeavoring to find out more and will report back as to exactly how it works and what it can do.

Overall though this is part of a push by companies like Autonomy, but also Attivio, Endeca, Exalead and some others into the search-based application market. The underlying premise of that market is database offloading; the idea of using a search engine rather than a relational database to sort and query information. It holds great promise, partly because it is the bridge between enterprise search and business intelligence but also because of the prospect of cost savings for customers as they can either freeze their investments in relational database licenses, reduce them, or even eliminate them.

Of course if the enterprise search licenses then get so expensive as to nullify the cost benefit, then customers will reject the idea, which is something of which search vendors need to be wary.

Users can apply to joint the beta program at a very non-Autonomy looking website.

I had a very interesting conversation with Attivio recently. Most of it is confidential and thus not enough for me to wrote a full 451 report on it yet (but I will as soon as we’re able to), but I can say that the startup founded mostly by former employees of FAST Search & Transfer that has a couple of public customers has some much more interesting news coming down the pike in coming months. It just announced one of these customers, here.

We talked about what is triggering the customer wins it is getting and one thing stood out; the key value prop Attivio has been focused on from the start, which is the conjoining of structured and unstructured data. But it’s not just about having access to both types (that was possibly years ago and anyway there’s myriad types between those two simplistic and oft-confused descriptions, but bear with me). And it’s not just the ability be able to query them both; that too has been done, but crucially with different tools in the past.

It’s the ability to switch between both based on what the user is doing – not based on what type of data is being queried – to do so without the user knowing and having a single API for access to both types. And that, say the Attivio team, is hitting home with customers.

Attivio is not the only company to be thinking like that of course. Attivio’s management’s alma mater FAST was doing this for a while as a product called Adaptive Information Warehouse. We wrote a report on it in January 2007 under the headline: FAST: everything you thought you knew about BI is wrong. And there are other companies both positioning their products this way and winning customers doing it. See our recent report on Exalead for another example and we’re sure IBM could build its customers just such a system if they paid enough.

The challenge with this idea is of course that for the past 30 years or so relational databases have been where the ‘important stuff’ has been stored and the multi-billion BI market grew on top of that as a way to access it. Database administrators rule(d) the roost as far as information management goes. Meanwhile enterprise search got relegated to a side room where it was all about finding documents and getting pages and pages of results returned to you. What Attivio, Exalead and a few other companies are moving towards is a convergence of the two; call it database offloading, unified information access; unified information intelligence or something similar.

We’re not seeing these vendors being dragged kicking and screaming towards this single-API-no-matter-what-the-data-is nirvana by their customers, however; it takes a fair amount of market education on the parts of the vendors and their partner to make it a reality. But given what we see happening here, we expect to hear alot more from the vendors mentioned here and others throughout 2009 and beyond as enterprise search morphs into something new.

I’ve gathered all my current thinking on potential M&A in enterprise search in a SectorIQ that we published earlier this week to our customers. In it, I look at four main potential targets plus a few other small ones and look at a few of the likely acquirers. (This is the way we write all our Sector IQs, btw and they’re a great way of getting a quick grasp on what might be coming down the pike in any particular sector of the IT industry)

Fortunately those of you that are not our customers (yet!) are able to read it via our arrangement with the New York Times DealBook section. Click here to see the NY Times posting or go here to go straight to the report – and while you’re there, sign up for a trial of our M&A KnowledgeBase, where we’ve been collecting details of every IT, internet and telecoms deal since the start of 2002!

Finally, a quick word about the headline. We like to have some fun here at 451 with these things and while I appreciate that this one might have been pushing things a little in terms of clearly explaining what the report was about, when else would I be able to use it?