Analytics help app developer conceive of new possibilities
By now, we all have realised the power of IoT, Mobile Apps, Big Data and Analytics. Now it’s time to use this power in every possible way for complete well being of everyone in the world. Let’s read this interesting article on Women Health Care Mobile Apps and Data Analytics.
By Paul Gillin.
Desperately trying to get pregnant? There’s an app for that.
Actually, there are many apps for that. Type “period tracker” into Google Play and you’ll uncover more than 100 programs that help women track their menstrual cycles and dispense all kinds of helpful advice about health, fertility and how to conceive.
One of the top-rated ones is called “Flo,” a free period calculator that tracks and predicts ovulation (above), the onset of a period and the days that are most conducive to conception. Flo uses a specialized calendar that lets users log stages in their cycle, schedule reminders and record mood and symptoms. It dispenses advice based upon its integrated rules engine. Thanks to new cloud infrastructure and a powerful backend analytics system being built by Minsk, Belarus-based InData Labs, Flo will soon do much more. (* Disclosure: This is a Wikibon case study sponsored by InData Labs.)
“Our goal is for Flo to be a simple tool and health companion,” says Marina Davydova, managing partner at OWHealth Inc., which developed the app. “However, behind the simplicity hides a sophisticated system that collects data on overall health, analyzes it with a help of artificial intelligence and provides recommendations for healthier living, so Flo can provide really valuable insights to its users.”
OWHealth has collected a vast quantity of data in the few months since the app launched in March, but until recently it lacked the means to do much with it. Flo was built on Parse.com, a mobile-back-end-as-a-service that Facebook Inc. acquired in 2013. Early this year Facebook unexpectedly announced that Parse would shut down in early 2017. That gave OWHealth enough runway for launch, but they immediately had to turn their attention to a migration strategy. That process, in turn, prompted Davydova and her team to think about ways to derive more value. If they were already switching hosting services, why not build out analytics capabilities at the same time?
OWHealth went searching for a partner that could make the migration seamless to its rapidly growing user base and also build out the advanced analytics. The partner needed to be able to tap into the richness of the information the app was collecting while also being attentive to the sensitive nature of the data.
“We are dealing with a huge amount of very intimate information, so it was crucial for us to provide both privacy and security,” Davydova said.
InData Labs is a big data consulting company that combines analytical expertise with project management and technology integration skills. “We chose InData Labs for their professional and sophisticated approach, from analyzing the business and technical aspects of our objectives to developing and implementing data strategy,” Davydova says.
A health trove
It turned out that the user data that Flo collected in just its first six months was a treasure trove of information that went far beyond reproductive advice. Mobile apps are fundamentally changing the field of medical research. In the past, studies on women’s health were limited by practical considerations to only a few thousand participants. Mobile apps like Flo have game-changing potential because users volunteer data that previously required expensive field research. The diaries of Flo’s 1.5 million users can yield fascinating correlations that uncover insights that help others. The app has become so much a part of their lives that many women update their calendars one or more times every week, even when they aren’t having or expecting a period.
The kind of analytics that OWHealth envisions wasn’t possible with the original Parse platform, which “doesn’t provide any ability to conduct analytics. It’s an easy to use but pretty database-centered kit” says Sergey Enin, delivery manager at InData Labs and solution architect of the project. InData designed a high availability lambda architecture using a tried-and-true object-relational engine that supports hundreds of sophisticated data manipulation and analytics routines. The backend was designed to support Flo’s evolution into a full-scale lifestyle advisor.
Changing backend infrastructure for a cloud application that already has more than one million active users is not unlike replacing the transmission of a car while speeding down the highway. “It’s a sophisticated process that involves many stakeholders,” Davydova said. “We could not afford to make any mistakes.”
Fortunately, InData has experience with such complexity. “After careful consideration of options available on the market we’ve chosen InData as the clear leader. Their product and execution is much better than competition,” Davydova says. The company created a hybrid batch/real-time process that enabled it to ease off the pressure on the production servers while creating batch extracts that could be moved in small batches. The entire migration took about three months, and Flo users didn’t notice a thing.
Analytics front and center
Now analytics takes center stage. The queries that InData is creating will yield insights that improve not only the quality of advice Flo provides but also the user experience. Constant monitoring of user behavior creates an ongoing feedback loop. “We want insights not just on how our users feel physically, but also on how they use our apps,” Davydova said. “Identifying those patterns will improve the consumer experience so Flo becomes the perfect woman’s health advisor while still being simple to use.”
The new intelligent system will aggregate mountains of data to tease out relationships that aren’t evident to human analysts. For example, pain symptoms can be compared against diet, exercise, sexual activity and mood to give advice on likely causes, or women can be advised on how to modify their diet and exercise routines during their period. OWHealth expects Flo to become more than just a diary; it will be an advocate for better health.
“The idea is to arm women with more information on their health to keep them confident and calm,” Davydova says. That advice can take many forms. Some people use Flo to track changes in their body. Others seek to better understand the relationship between patterns in their cycle and other health data. “People contact us regularly with stories about how our app has helped them to become more aware of their health in general,” Davydova says.
As InData extracts greater amounts of intelligence from OWHealth’s database, the app developer plans to sell subscription services that incorporate data from external sources, provide detailed analytics and personalize recommendations. For now, the app remains free and the rewards are principally in the glowing user feedback that streams in. “Users like the functionality, design and accuracy of our apps,” Davydova says. “Couples reach out to tell us that they finally conceived thanks to our help.”
Original post. Reposted with permission.
Bio: Paul Gillin is the Senior Editor for Wikibon’s micro-analysis team. He is the author of five books and more than 300 articles on the topic of social media and digital marketing. Gillin has 23 years experience in tech journalism, including his time as founding editor-in-chief of B2B technology publisher TechTarget as well as editor-in-chief and executive editor of the technology weekly Computerworld. He is a Senior Research Fellow at the Society for New Communications Research and a member of the Procter & Gamble Digital Advisory Board.
- Magpie Health Analytics: Health Data Scientist
- Predictive Analytics World: Hot Topics in Analytics for Healthcare
- 7 Big Data Steps in Health Science