The HarperDB Hackathon concluded on Monday with five stellar finalists but only one winner. Congratulations to Team Pangongso on your incredible Healthcare Portal example using HarperDB. The sole member of Pangongso, Ashwini Kumar, is an accomplished technical architect out of Boston working in finance. The goal of Ashwini’s healthcare portal was very similar to that of HarperDB, to simplify a big data architecture without sacrificing quality and capabilities of the product. An example that not only uses HarperDB in a new and interesting way, Ashwini’s portal also makes the patient’s life easier and with an almost near perfect score, we were excited to do a feature on this project.  

The Challenge Pangongso Took On: 

Today, healthcare records are outdated and siloed throughout many different systems. Every time you go to a different doctor, a new or existing file is updated for you at that location. This file stays stagnant. How many doctors’ offices provide you with a login for a portal to view your data? This fragmentation can be frustrating for both patients and doctors.  Aggregating these fragments from disparate sources so as to facilitate effective collaboration among various stakeholders responsible for the physical and mental well being of a patient has become a significant challenge.  Security and privacy is of primary concern for such an aggregation to succeed.  

For effective healthcare, any system supporting it needs to react with extreme agility to changes in terms of data acquisition.  Our ability to rapidly change data collection forms and quickly publish health related alerts have a direct bearing on the quality and cost of care.   

As Ashwini states in his presentation about his portal, healthcare is a collaboration between a few key stakeholders. But those key stakeholders including your various doctors, family members, nurses, mental health therapists, and even fitness instructors or nutritionists, don’t have a way to communicate. You might not even know how to view your own data without specifically asking for that information. How many times do you fill out the same general patient information form in a year. Wouldn’t it be nice if instead of waiting on a call weeks after getting lab tests, if you could get a push notification the minute your results are in? 

Ashwini, asked the question, how can we use an HTAP database to solve this problem? 

Solution: One Portal for All of Single Patient’s Data both SQL & NoSQL

Using HarperDB and Liferay Ashwini came up with a solution.  Using HarperDB as the foundation, he found a means of populating our database with data from various healthcare monitoring systems by utilizing the flexibility of HarperDB’s Dynamic Schema. Various records and patient information would then invoke rules and push real time alerts to a user interface so that patients and their key stakeholder can view their data and interact with their data in real time. Your health documents would be at your fingertips and HarperDB acts as the gateway and repository for external sensors to send these alerts and notifications.  

We have already established that HarperDB is an IoT database and it’s built for these scenarios such as edge computing where critical updates to data need to happen in REAL real-time. Being able to ingest both unstructured SQL data and ACID compliant NoSQL data allows for efficient ingestion of both the static and highly dynamic data that is collected throughout a patient’s lifetime. The result is a collaborative portal that meets the need of every different stakeholder’s dynamic data sets that they provide.  

“HarperDB’s design has enabled innovation at the grass roots level. This can be a catalyst to transform the healthcare experience for all stakeholders”- Ashwini Kumar, Team Pangongso 

The Technology: Using HarperDB as a IOT Database for Healthcare

 Ashwini really took a deep dive with HarperDB and that was pretty exciting for the HarperDB engineering team.  Kyle our CTO was particularly impressed that Ashwini used the direct file system access that HarperDB provides to create his own triggers and workflow using an awesome library called chokidar.  This allowed Ashwini to extend the capability of HarperDB to handle more complex application logic.  Incidentally this is something on our roadmap that our dev team is currently working on; however, seeing a live solution from the community using HarperDB was awesome.  He also broke the project into two components, one for the sensor side which heavily utilized JSON and the NoSQL interfaces of HarperDB, while more heavily utilizing traditional SQL on the server side.   

Finally, another really exciting part of Ashwini’s project, that shows the community that is springing up around HarperDB was that the solution utilizes a library, harperdb-connect, previously built by one of the other Hackathon contestants, Ethan Arrowood.   Its pretty neat to see the community beginning to build libraries and solution around HarperDB organically and imagining where these new projects will take the HarperDB community.  

If you would like to learn more about this developer example and more, check out our full library.