The Data Guild Labs

Guild Labs is a collaborative space where we develop and test data products to address significant market needs. When we identify a promising market opportunity, we assemble a core team representing product, data, engineering, and design to develop a prototype that’s ready for market validation. From there, we iterate and refine with target customers, bringing resources to bear as needed to get to market fit. The combined community of Guild members, strategic partners, and customers offers us an excellent opportunity to build data products that matter while maximizing our chances for success. We’re developing a significant IP portfolio which we either license to strategic partners or spin out into stand-alone companies.

Below are some examples of current Labs data products:

Faer

iphone-6-hand-demo

Some 80% of healthcare costs in the U.S. today are attributable to chronic disease. Faer is a mobile-based adaptive assistant that helps those patients live better lives. It uses artificial intelligence to deliver personalized support, access to services, and connections to a care team, learning incrementally how to better serve patients in the future and reduce healthcare costs. For community healthcare workers, nurses, and doctors, it also provides a window into patient progress, helping the care team concentrate scarce resources on the patients who most need it. Status: Piloting with the Center for Medicare and Medicaid Services

 

Patterna Health

sensor

A routine challenge for health care administrators is balancing the demands of training and policy enforcement with the consistent delivery of high quality care. Often they are forced to make operational tradeoffs without a clear understanding of current and emerging issues in the patient care environment. Patterna Health addresses this deficiency using low-cost, ultra-low-power proximity sensors and a simple, actionable analytics delivered via mobile or secure web. By integrating patterns in human mobility, relative proximity and co-location with hospital surveillance information, Patterna highlights potential risk patterns earlier giving administrators better information to maximize patient safety and quality of care. Status: Prototyping

Motion

network

How do you create persuasive, tailored messaging that motivates people to act?  Encouraging people to change behavior in healthcare and public awareness settings requires a nuanced approach.  A key challenge is using heterogeneous data to construct candidate cohorts, match those with tailored messaging, and learn from the results to improve messaging effectiveness over time. Motion is a system built on an extensible architecture that can construct customized behavioral messaging over large-scale populations and learn from repeated experiments as individuals take action. Status: Spun out into Motiva.ai

 

Tripwire

eye

Increasingly our globalized world is dependent on complex systems that are optimized for efficiency and yet susceptible to cascading failures. Without additional automation, humans are unable to comprehend the state of these systems and identify potential abnormalities. We are developing Tripwire to monitor heterogeneous data sources for malicious behavior that could threaten the stability of the global financial markets. With a tailored workflow supporting efficient exploration and incremental learning based on user feedback, the goal is to support more timely discovery, diagnosis, and intervention. Status: Piloting

 

IoTa Efficiency

iota

Many operational industrial systems were designed years ago with modest sensing capabilities to satisfy requirements for system control. Attempts to optimize the efficiency of these systems are often stymied by high latency and poor quality in the existing measurements collected. The purpose of IoTa is to augment traditional systems with wireless, low-power sensors for the explicit purpose of real- and near-time optimization and control. IoTa enables a high-quality collect-learn-recommend loop required for dynamic optimization in energy, operations, and manufacturing scenarios. Status: In development