How to add analytics to your application development pipeline
Discover a new, easier way to enhance Python applications with an open analytics platform
By Alison Bolen, SAS Insights Editor
Let’s say you manage a team of smart application developers. They build engaging experiences for your customers on the web and through mobile apps. As a result, your web traffic has doubled in the last year, and your app download rates are tripling every six months.
Now you’ve challenged your team to take it to the next level and add some analytical intelligence to your web and mobile properties. You want to provide visitors with a unique, individualized experience based on where they’re located and how they’ve interacted with you in the past.
Adding that type of insight into your apps requires advanced analytics on some pretty large data, so your developers will need to embed analytical capabilities directly into the apps they’re building.
Then, rather than figuring out how to consume assets coming from the data scientists, app developers can now focus on building apps that look great and work perfectly.
Cloud and Platform Technologies
To accomplish this, you reach out to your data scientist. She builds an analytics model that analyzes customer browsing history and location data, and sends it to the application development team. Now the application developers have to recode that model to run as part of the app.
What if there were an easier way? What if the data scientist could build a model, right-click on it and make it available easily – say as a REST web service? The application developer can now simply reference the model directly from a REST call within the app and never worry about it again. This is can be accomplished easily with an open analytics platform.
“Then, rather than figuring out how to consume assets coming from the data scientists, app developers can now focus on building apps that look great and work perfectly,” says Frost.
What are the overall benefits of an open analytics platform for your software developers?
- Application developers and data scientists can focus on what they do best without wasting a lot of time trying to integrate technologies.
- Time to market for apps is shorter because models built by data scientists and made sharable via REST are instantly available for use and reference by app developers.
- The validation process for rolling out apps becomes faster since model building, tracking and maintenance are all handled independently of the application engineering process.
- You reduce the risk of broken code when changes happen in app development.
- Analytics capabilities are uniformly enabled, not stitched together from different packages, analytics disciplines or coding languages.
- Scalable analytics are backed by processing power, requiring no code changes as data grows in volume or speed, since the REST calls are made in an infrastructure optimized for analytics.
- Governed models and automated decisions become their own entities, and you can define ongoing monitoring and alerts for tracking deployed result performance.
- A centralized pool of models makes it easy to swap out champions with challengers without changing application code.
- Developers can individually call the same advanced analytics capabilities used by data scientists to build models from within any application.
“Overall, an open analytics platform provides substantial productivity improvements for everyone. The applications team can create more apps faster, and the modeling team can focus on innovative algorithms,” says Frost. “Everyone can be 100 percent focused on their job, instead of taking time away from what they do best, to manage the integration between modeling and application building.”