GVK EMRI provides rapid medical care and emergency transport to nearly 700 million people in 17 Indian states and
union territories per year. Over the past decade, GVK-EMRI has provided specialized training to over 10,000 EMT’s,
who provide care to an average of 22,000 patients in India every day. The data from these activities has been
organized and shared with Stanford Hospitals as a part of an on-going relationship to improve care.
Datorium has worked over several years with Stanford to deliver a unique AI-driven approach to healthcare, identifying opportunities to improve emergency care to the one billion lives covered by Indian emergency medicine services. Using the Octain platform, we can identify interventions specific to certain patient segments to improve outcomes and continue to learn over time.
In addition to acute care, we’ve developed methods to parse electronic health records (EHRs) using
natural language processing, developed novel data products from health wearables and worked with the
largest US healthcare claims processors to identify opportunities to improve efficiency and profitability.
Fraud accounts for a loss of nearly 12% of all cash transactions in US retail, and even higher overseas. In this engagement, Datorium worked with one of the largest US retail beverage provides to reduce fraud by predicting the probability of fraud at multiple levels. By ranking our score between zero and 1, analysts are able to sort through the noise and focus on the most egregious cases of fraud, saving millions in lost sales directly and indirectly (through the network effect of intervention).
The Gates Foundation in partnership with Rockefeller Philanthropy sponsored Datorium to study mobile financial services retention in sub-Saharan Africa (Kenya, Tanzania, Uganda). In this engagement, we were able to identify key drivers for long-term adoption of mobile financial services, as well as influencers of deeper engagement. The methodology developed in this study became the basis for our Octain Pulse churn platform– understanding how users’ behavior changes over time is critical to understanding their motives and how to
decrease churn, improve engagement and improve retention.
Buildings produce more CO2 and carbon gasses than any other source. Within buildings, HVAC systems consume the most electricity, and within HVAC, air conditioning is the least efficient. In this engagement, the Datorium team was
able to build and deploy a novel optimization system, which 1) predicted future load 2) developed an optimized sequence of equipment to address the load and 3) deploy this recommendation to achieve cost savings and CO2 reduction. The system is now in use by dozens of large organizations in the US through our partnership with Optimum Energy.
Global supply chains present an AI challenge in that data is often disparate, high-latency, messy and sparse. In this project, Datorium worked directly with HP’s global supply chain management team to develop predictions of unit demand, improving financial forecasting and streamlining manufacturing to prevent overage and increase profits.
Net Promoter Score (NPS) is often used as a customer health index and provides predictive power towards future purchasing and revenue. However, what to do when your data is limited to a small subset of your customer population? In this study, Datorium worked with senior managers at VMware to develop a system to predict NPS for the entire customer base. Using Octain, the team iterated over many internal and external data sources to develop a highly-accurate and highly-explainable model, providing influence and providing insight never before developed about the diverse global enterprise customer base.
In a separate study, Datorium worked with F5 to develop a lead scoring model with specific recommendations on a per-customer basis. This "vehicle and cargo" methdology helped the team to understand not only who to talk to, but what to talk about and how to reach them.