Durham, NC, startup ArtusLabs this week said that it has closed a $2.6 million Series A round of venture-capital funding to help it develop and market data-management software for the life-science sector.
Robin Smith, co-founder and CEO of ArtusLabs, told BioInform that the 10-person firm plans to use the funding to support new product development and expand its marketing efforts for its flagship product, a software system for organizing, managing, and sharing unstructured scientific data called Ensemble for Life Sciences.
At the American Chemical Society meeting in New Orleans this week, the company launched the system, along with its first component, a chemical database called Ensemble Exchange.
Smith and ArtusLabs co-founder Brian Ballard, who serves as CTO, are not newcomers to the life science informatics market: In 2000 they founded Synthematix, a provider of electronic laboratory notebooks. The company was acquired by Symyx Technologies in 2005 for $13 million.
Smith said that every one of ArtusLabs’ employees is a former Synthematix staffer, a feature of the company that he described as “unique” for a startup. “We’ve all worked together for seven years,” he said, “so it’s easy for us to build products and address challenges in the marketplace.”
ArtusLabs’ staffers aren’t the only familiar faces at the firm. Hatteras Venture Partners, which with Southern Capitol Ventures led the $2.6 million round, was also the founding investor in Synthematix, and Clay Thorp, general partner at Hatteras, was the first CEO of the original firm.
“We had worked with these people successfully before, we trusted them, we could finish each others’ sentences,” Thorp said, adding that the sale to Symyx netted his firm around a five- to six-fold return on its investment.
Given this prior success, he said, “we were pretty open-minded to pretty much anything [Smith] was doing.”
With their latest venture, Smith and Ballard have moved well beyond electronic lab notebooks into what Smith described as “decision management” tools for research organizations.
This opportunity “is much bigger than what we had explored with Synthematix,” Thorp said. “And it’s focused on a much bigger problem, which is, ’How does a research-oriented company learn from what it’s done, and how does it share that information in a community of learning and innovation?’”
Ensemble for Life Sciences aggregates into a central database all the unstructured data within a research organization, including biological and chemical data files, research reports, spreadsheets, PowerPoint slide decks, and the like, and then uses Web 2.0 technologies inspired by websites like Flickr and del.icio.us to present that data in a way that is easily managed.
“We had worked with these people successfully before, we trusted them, we could finish each others’ sentences.”
“We’ve been looking at this unstructured data problem and the one market segment that’s really figured it out is the consumer side of social networking business,” he said. “They’ve really figured out a way to apply context to things that are just kind of everywhere, and the chaos that happens with consumers interacting on the website.”
Smith said that ArtusLabs adopted many of the same technologies and architectural techniques that these sites use, including features like tag clouds and Ajax-style programming. “These techniques apply directly to the business environment and the science environment because there is tons of unstructured data that is really complex and chaotic and you really don’t want to make the user conform to a certain way of doing things, because then they can’t actually be free to innovate,” he said.
The key application for the system, Smith said, is capturing a company’s decision-making process, which often relies on unstructured data that employees bring to meetings, either by “carrying it in a folder, or carrying it around in their heads.”
The problem, he said, is that over time, “people forget what that unstructured data was that they actually used to support those decisions.”
This situation is “amplified” in pharma and biotech companies, Smith said, because of the industry’s decade-plus product-development lead time, which is more often than not accompanied by numerous staff changes. By the time a compound is ready for clinical trials, there may be no record of why it was even selected for development.
Smith said that the system can be configured to collect data from multiple sources, including directly from instruments. The software also has a “mass importer” feature that can retrieve data at routine intervals from various sources.
“We’ve taken the same approach that companies like Google did,” he said. “Google doesn’t go out and search multiple websites when you do a query. It searches an internal index they’ve built. So what we do is kind of the same thing. We take the documents and assets and unstructured data, and instead of tying it together or integrating it or federating it, we centralize it — we pull it into our system. And then we can apply all kinds of state-of-the art algorithms and computational techniques to that data, and it’s fast.”
There are also several deployment models for the system. ArtusLabs will host it for smaller companies on its own servers, or install a “black-box” system that includes a server with several terabytes of storage arrays. The user interface is browser based, “so there’s no software to install on desktops and manage,” Smith said.
ArtusLabs has signed three undisclosed alpha customers for Ensemble for Life Sciences: a small pharmaceutical firm, a biotech, and an academic institute.
The academic institute is using it “as a collaborative tool to share data with other academic institutes and their collaborators,” Smith said, while the pharma is using it to aggregate a large amount of legacy data into a single system for searching. “We saw them looking for some of their compound numbers and finding one in a slide deck that was from a year before they even thought they had started working on that project,” he said.
The biotech, meantime, “is kind of a hybrid of both of those. They’re using it to manage their historic content and also use that in a collaborative sense with different groups that are working on various programs.”
Smith said that the primary competition for the company’s technology is in-house systems that companies have developed in an attempt to federate data from multiple sources.
“I think the challenge for us is getting [customers into] that mindset of, ‘You don’t need to have five systems. You can have one system now. You can start not having to support the others,’” he said. “When they start thinking about that, it becomes sort of a no-brainer.”