Top Healthcare Provider Derives Actionable Medical Insights from Terabytes of Clinical Data Using Pachyderm’s Scalable, Data-Driven Machine Learning Pipelines
Customer Company Size
Large Corporate
Region
- America
Country
- United States
Product
- Pachyderm
Tech Stack
- Artificial Intelligence
- Machine Learning
Implementation Scale
- Enterprise-wide Deployment
Impact Metrics
- Productivity Improvements
- Cost Savings
Technology Category
- Analytics & Modeling - Machine Learning
- Analytics & Modeling - Big Data Analytics
Applicable Industries
- Healthcare & Hospitals
Applicable Functions
- Product Research & Development
Use Cases
- Predictive Maintenance
- Remote Asset Management
Services
- Data Science Services
About The Customer
The customer is one of the top for-profit managed healthcare providers in the U.S., with affiliate plans covering one in eight Americans for medical care. They have a mission to be the most innovative, valuable, and inclusive partner in health benefits. The company has a dedicated AI team who are looking to leverage cutting edge AI to harvest long term insights and make much more detailed health predictions from claims and electronic health record data. The data store is massive, with more than 50 terabytes of data covering the company’s tens of millions of members across the U.S. They are mining this data to determine treatment efficacy based on past outcomes given particular patient characteristics.
The Challenge
One of the top for-profit managed healthcare providers in the U.S., with affiliate plans covering one in eight Americans for medical care, was looking to leverage artificial intelligence (AI) to harvest long-term insights and make much more detailed health predictions from claims and electronic health record data. The data store is massive, with more than 50 terabytes of data covering the company’s tens of millions of members across the U.S. They were mining this data to determine treatment efficacy based on past outcomes given particular patient characteristics. However, getting these potential insights into the hands of healthcare providers was a challenge. It’s one thing to have small scale implementations working in a lab, it’s another to deliver machine learning at scale. When the engineering lead joined the AI team, they had a very complicated data delivery pipeline based on Apache Airflow. While it worked, it wouldn’t scale beyond a single pipeline or container instance at a time.
The Solution
The healthcare provider turned to Pachyderm, a data layer that allows machine learning teams to productionize and scale their ML lifecycle. With Pachyderm’s industry-leading data versioning, pipelines, and lineage, teams gain data-driven automation, petabyte scalability, and end-to-end reproducibility. Pachyderm delivered the parallelism and data handling required to efficiently scale the AI team’s ML processing. Importantly, while the company had millions of patient records, only a small subset were relevant at any given time, and Pachyderm’s incrementality saved significant time, money, and resources by only processing the subset of data that had changed, rather than the entire patient universe. With Pachyderm, the team was able to arbitrarily partition table data to only capture events for a single member – effectively creating individual member objects that encapsulate all the events for a particular member. Pachyderm not only processed these records in parallel, it also automatically processed only those containing new information, increasing both scale and speed while reducing costs.
Operational Impact
Quantitative Benefit
Case Study missing?
Start adding your own!
Register with your work email and create a new case study profile for your business.
Related Case Studies.
Case Study
Hospital Inventory Management
The hospital supply chain team is responsible for ensuring that the right medical supplies are readily available to clinicians when and where needed, and to do so in the most efficient manner possible. However, many of the systems and processes in use at the cancer center for supply chain management were not best suited to support these goals. Barcoding technology, a commonly used method for inventory management of medical supplies, is labor intensive, time consuming, does not provide real-time visibility into inventory levels and can be prone to error. Consequently, the lack of accurate and real-time visibility into inventory levels across multiple supply rooms in multiple hospital facilities creates additional inefficiency in the system causing over-ordering, hoarding, and wasted supplies. Other sources of waste and cost were also identified as candidates for improvement. Existing systems and processes did not provide adequate security for high-cost inventory within the hospital, which was another driver of cost. A lack of visibility into expiration dates for supplies resulted in supplies being wasted due to past expiry dates. Storage of supplies was also a key consideration given the location of the cancer center’s facilities in a dense urban setting, where space is always at a premium. In order to address the challenges outlined above, the hospital sought a solution that would provide real-time inventory information with high levels of accuracy, reduce the level of manual effort required and enable data driven decision making to ensure that the right supplies were readily available to clinicians in the right location at the right time.
Case Study
Gas Pipeline Monitoring System for Hospitals
This system integrator focuses on providing centralized gas pipeline monitoring systems for hospitals. The service they provide makes it possible for hospitals to reduce both maintenance and labor costs. Since hospitals may not have an existing network suitable for this type of system, GPRS communication provides an easy and ready-to-use solution for remote, distributed monitoring systems System Requirements - GPRS communication - Seamless connection with SCADA software - Simple, front-end control capability - Expandable I/O channels - Combine AI, DI, and DO channels
Case Study
Driving Digital Transformations for Vitro Diagnostic Medical Devices
Diagnostic devices play a vital role in helping to improve healthcare delivery. In fact, an estimated 60 percent of the world’s medical decisions are made with support from in vitrodiagnostics (IVD) solutions, such as those provided by Roche Diagnostics, an industry leader. As the demand for medical diagnostic services grows rapidly in hospitals and clinics across China, so does the market for IVD solutions. In addition, the typically high cost of these diagnostic devices means that comprehensive post-sales services are needed. Wanteed to improve three portions of thr IVD:1. Remotely monitor and manage IVD devices as fixed assets.2. Optimizing device availability with predictive maintenance.3. Recommending the best IVD solution for a customer’s needs.
Case Study
HaemoCloud Global Blood Management System
1) Deliver a connected digital product system to protect and increase the differentiated value of Haemonetics blood and plasma solutions. 2) Improve patient outcomes by increasing the efficiency of blood supply flows. 3) Navigate and satisfy a complex web of global regulatory compliance requirements. 4) Reduce costly and labor-intensive maintenance procedures.
Case Study
Harnessing real-time data to give a holistic picture of patient health
Every day, vast quantities of data are collected about patients as they pass through health service organizations—from operational data such as treatment history and medications to physiological data captured by medical devices. The insights hidden within this treasure trove of data can be used to support more personalized treatments, more accurate diagnosis and more advanced preparative care. But since the information is generated faster than most organizations can consume it, unlocking the power of this big data can be a struggle. This type of predictive approach not only improves patient care—it also helps to reduce costs, because in the healthcare industry, prevention is almost always more cost-effective than treatment. However, collecting, analyzing and presenting these data-streams in a way that clinicians can easily understand can pose a significant technical challenge.
