While speaking about the challenge of harvesting value from massive volumes of archival data, Walt mentioned the continuous stream of new unstructured data that includes emails, social media posts, forms, business processes, and more.

“Something on the order of 80% of all of the data in a typical organization falls into this unstructured category, much of which is in the form of language,” Walt said. “How do you try to unlock the value of that language data, so that you’re getting the right information to the right people to help them make faster, smarter, more consistent decisions?”

To generate domain-relevant data that is practical and purposeful, Walt recommended engaging with a solutions integrator that is experienced in handling voluminous, complex, unstructured data.

“And I think that’s in particular where Titan Technologies excels,” Walt said. “They bring the solutions to bear that are going to make a difference in the ability of an organization to use language data for faster, smarter, and more consistent decisions.”

 

Creating solutions today that bring more benefits tomorrow

Creating new technology solutions that are practical and purposeful starts with understanding a customer’s existing investments in data management and search infrastructure, and creating ways to augment — not necessarily replace — that installation to gain significantly valuable functionality.

Brian noted that any proposed data management solution must help organizations put their language data in context, while still being mindful of the specific requirements of each organization. He also mentioned how the flexibility of cloud technology is opening up so many new customizable data management options for customers.

When briefing organizations about Compendia, Titan Technologies’ modern insight engine, Brian emphasizes that Titan’s goal for each customer is not just to enable them to have better access to their archived data, but also to develop ways to leverage the technology even further.

“The easiest way to think about it is related to data search, where you’re looking for something specific,” Brian said. “We can help you do that, but there are also downstream products that you can get out of this. And how we, as Titan, expose that to you as a customer, makes a difference in terms of what you could ultimately do with it. So we’re practical in terms of producing a product for you. At the same time we’re also producing it and storing it in a way that can serve your current business needs, and put you in position to do things with it in the future as well.”

As an example of how natural language understanding technology can open up new capabilities for government agencies, Brian cited Titan Technologies’ deployment of Compendia for the Los Alamos National Laboratory’s National Security Research Center. Those researchers might need to find information about obscure experiments conducted as far back as the 1930s, so Compendia’s ability to understand complex search queries and extract the right data is crucial.

“There is incredibly valuable information that needs to be more easily surfaced and referenced within the context of those experiments,” Brian said. “Discovering and understanding the parameters [of the experiments] and the outcomes makes it so much easier for the researchers to quickly benefit from and build on that important science.”

 

Taking the first steps to better data management

For organizations (either private enterprises or government agencies) taking their first steps into data management technology, Walt suggested that they first think about what they are most focused on that is core to their mission.

He then advised that these organizations analyze what kinds of data are most relevant to their domain, then determine questions that could be answered with the data that they have or the data they want to capture.

“And then from there, I think the use cases start to emerge,” Walt said. “We use a framework of analyze, understand, and accelerate. And that’s applied to language in particular. Analyze is when the complexity is high. Accuracy is important and mistakes are costly. And that typically requires an expert. And by applying the natural language understanding capability that we have, we can embed the knowledge of that expert. Then we can enable everybody to have a new, consistent level of understanding of documents, and then take it from there.”

Brian noted that the best way to start a data management project is to have a very frank and complete discussion upfront about what you want to achieve and how that can help your organization. By starting small, achieving objectives, and learning from the experience, the project can follow an effective, agile, iterative building process.

“We are a team of problem solvers,” Brian said of his Titan Technologies teammates. “We’ve seen a lot of different things. We’re very interested in having that kind of upfront discussion about what you want to achieve, and how does that work. And then working with our partners, like expert.ai to sit down and craft out a solution that makes sense.”