CEO Bill Coleman - the "B" in BEA - says there's one more chance to save Cassatt, the ahead-of-its-time data center automation company that he started six years ago after leaving BEA where he had been its first CEO.

He should know by Friday. He's got his fingers crossed, he said.

He thought he had a deal to sell the company months ago only to have it wash out.

Some of leading lights of the industry have turned thumbs down on buying the company. Bill puts it down to NIH as with HP or in the case of Amazon, Google and Citrix their different, more limited business models.

If this last ditch effort doesn't pan out this week then Bill said he will put Cassett's IP up for sale and turn off the lights. The IP, as sexy as it might be, is unlikely to fetch the $100 million Cassatt has burned through since it kicked off.

If the deal does go through then Bill will be signing a couple of customer contracts worth a few million dollars almost immediately.

They're sitting there waiting to be signed, he said, but he can't do it until he knows Cassatt has a future.

One contract is with a government agency interested in Cassatt's disaster recovery skills. Bill says it can fail over and restart an entire data center in 45 minutes without any human intervention, something the government should be interested in.

But according to Bill, it's that kind of thinking that essentially doomed Cassatt. There's such a thing as being too disruptive. Cassatt's technology would have sidelined too many computer priests paid not to save their companies 50% of their IT costs but simply to keep what exists up, running and available "5Nines."

It didn't help that the push to do what was essentially a "heart/lung transplant," he said, was coming from a little no-name company despite all the industry prima donnas that inhabited the place, another mistake Coleman confesses to.

Cassatt wasn't the tight ship BEA was when it started because Coleman didn't have as versatile a number two as he had in Ed Scott, the "E" in BEA, and he had this chance to hire all these heavy hitters.

After Cassatt finally figured out that doing things its way wasn't working - God knows it was warned often enough but when have men ever listened - it tried making its technology into less threatening baby steps like this disaster recovery business, something it could have peddled initially but didn't. That wasn't the plan.

Then the cloud chatter started last year - raising Cassatt's hopes.

It can after all make Amazon EC2 resources a virtual part of an internal data center with no change in applications and less risk to data - but then the recession hit and users' pocketbooks snapped tight shut but not before Cassatt's pipeline got "really big," Bill says, claiming that that big pipeline is still there.

And it hasn't disappeared, he says, even if Cassatt's two biggest deals fell though on that day in January when that plane hitched in the Hudson River. Business had stopped closing around Christmas.

Bill says he "vastly underestimated the social and cultural challenge of Cassatt." People only think they want automation. The cost savings and proven ROI didn't justify the disruptive change it demanded in policies, procedures, processes and skill sets.

That's why he only has a dozen customers although everybody's interested - interested enough to listen and maybe do a pilot - but production is another matter.

He also discovered that people want a simple cloud and aren't really thinking about supporting legacy code.

They're enamored of virtualization, which, he says, adds complexity, increases OPEX, only marginally lowers server costs and only addresses scale-up, not scale-out and clustering.

Anyway, Bill wouldn't say who his knight in shining armor was. It might jinx the deal, he said.

This potential acquirer is supposed to want the company, its 52 people who have been paid up and on furlough since April 15 and its business plan the way it is and would reportedly pursue Cassatt's next-generation product code named Skynet after the fictional self-aware AI computer system in the Arnold Schwarzewnegger movie "The Terminator." Coleman's Skynet is supposed to blur the lines between data centers.