Roth’s Startup NexLP Develops Automated Document Understanding
If you were to give most people a single email message and ask them to report back with a brief report of the contents and a list of the key players in the message, most anyone would be able to do it fairly quickly. If you gave them, say, 10 email messages and sufficient extra time, they could probably do the same again.
If you were to hand 10,000 email messages, things would likely start to break down.
But this is just the type of scenario that many corporations are finding themselves in, particularly where legal aspects are involved. Legal proceedings can now require the review of thousands of submitted documents, including memos and email messages. The only way to get through these and determine what documents may be of relevance was to hire many, many people to skim and make judgments on the applicability of document.
Now, however, a startup company—NexLP—co -founded by CS Professor Dan Roth, an expert in natural language processing,is developing an automation process for reviewing the vast amounts of documents that can be found in legal proceedings.
“If you give me 200K messages, which you can never read, what we can do immediately is tell you who are the key entities that are involved in this huge collection of data,” said Roth. “Who are the people that are mentioned the most, what they talk about and with whom. What are the locations, organizations, and other entities and topics that are mentioned the most? We can slice it different ways.”
“Basically, the core technology of NexLP is the tools that we’ve been developing here over the last 15 years or so,” said Roth. NexLP has licensed Roth’s technology from the University of Illinois to develop this into a marketable product.
A big boost for NexLP came when it was named one of the Techstars Chicago class of 2014. Each year 10 companies are selected for the Chicago Techstars summer program of mentoring, business coaching, and networking. In addition, the Techstars program provides funding, legal services, travel assistance, cloud services, and several other incentives.
The companies in the Chicago program receive office space in the 1871 facility for the duration of the Techstars program.
“The key thing is they give companies intensive coaching in how to run a company, how to build a company, how to interact with venture capitalists,” said Roth.
Being selected to Techstars is prestigious for up and coming companies, and many keep reference to that affiliation. The Techstars website indicates that companies that have taken part in the program have gone on to raise an average of $1.6 million in outside capital.
In August, the companies give a final demo of their product. It is at that point that companies could solicit additional venture capital assistance. But Roth thinks NexLP may not need that. “We already have sales,” he said. “We have already sold the product to a couple of companies. And we have pilot installations in maybe ten companies already.”
One reason that NexLP may not need additional venture capital funding is that the other founders of the company have experience working within the legal community. Roth’s fellow co-founders Ye Chen, Jay Leib, Kit Mackie, and Alan Rosen bring years of IT experience in the legal market to bear on the company.
“One reason I found this an interesting opportunity is that several of the people who founded this company with me are extremely knowledgeable about this [legal] area,” Roth explained. “They know the domain well. They know what people care about. They know the community.”
This knowledge of the legal community makes it easier to penetrate the market. And the company has been getting some press from such outlets as the Law Technology News. This could definitely help them with getting their product into the hands of the legal community. “This community is a relatively conservative community,” Roth said. “They are not known for adopting quickly high-tech things. We have to figure out how to convince them.”
And once they have settled into the legal market, Roth sees other avenues for developing the technology. “As we move on, there’s a lot of areas in big data, where companies have a lot of unstructured data, and they have issues of compliance, data security of different sorts, inside information, even internal spying, patents, and stuff like this. They need to know what’s in their data collection. “
As long as there are increasing amounts of data, Roth sees potential for the company to move into this market.