Have you heard about the Detroit couple convicted of stealing hybrid car secrets from GM and trying to sell them to China?

Or, have you read one of many stories about executives who jump from one company to another and try to illegally take trade secrets with them?

How about when federal regulatory agencies accuse big corporations of illegal practices?

That’s when the legal headaches begin, starting with who’s going to go through hundreds, thousands or millions of e-documents to find out which ones might be relevant to the case.

This process is called e-discovery or data mining. Some tech companies are striking gold with data mining, including Nashville-based DSi (formerly Document Solutions Inc.)

“The trend in the market that’s coming,” says independent computer security expert Bryant Tow, [is] “when a company has experienced corporate espionage or a breech, they call DSi to come in and do work.

“Sometimes corporate espionage is just attackers coming in and looking for one, two or three e-mails, looking for planned mergers or announcements on earnings so they can speculate or change negotiations,” Tow adds.

DSi gets access to every staffer’s computer work space, e-files, e-mails, even posts to social media.

A gigabyte of data is something like 10,000 pages of documents, and each staffer can generate three to 15 gigabytes.

Multiply that by 100 employees, or even 1,000 employees, and you’ve got millions of e-pages to review.

And you might be looking for just a few relevant documents.

In its searches, DSi uses filters and special software to find keywords.

“They have all kinds of fancy tools,” says Chad Schmidt, the marketing/communications director for Counsel on Call, a Brentwood company that uses DSi’s services to help corporate clients. “They put a flash drive in a hard drive, and it takes five minutes. They do that for everyone involved and they compile it and cull it.”

Adds John Burchfield, DSi’s vice president of business development: “We find the needle in the haystack.”

Or, more accurately, DSi clears away most of the haystack so the company and its law firm can better look for the needles.

“DSi is really good at working with our attorneys and culling it down to where 90 percent isn’t relevant, and attorneys only have to look at the 10 percent,” Schmidt explains.

“A million down to 100,000 documents, reducing 90 percent, is a really good result,” Schmidt says.

“The industry average is much less than that.”

And the relevant information is almost always there in the remaining documents, he adds.

“There’s no such thing as something [DSi] can’t find,” Schmidt says. “If it’s been entered into something, they find it.”

The savings for the company comes from not having to hire $300-an-hour lawyers to go through all the documents.

Schmidt estimates an average case might cost less than $5,000 in charges from DSi, but those costs can grow when all the culled data has to be stored. If it takes a year or more for a case to go to trial, costs rise significantly.

“That’s where it gets expensive,” he says.

But that’s all relative, he adds: “Large corporations spend tens of millions a year… to protect a brand.”