Apigee Redefines Big Data Analytics in the App Economy
Insights Empowers Organizations to Make Better Business Decisions by Analyzing Data from APIs, Apps, Social and Mobile Ecosystems – with Context
As organizations increasingly interact with customers, developers and partners through mobile, social and other app experiences, the data it needs to understand is expanding beyond the enterprise core. For example, when a company exposes its services through an open API, important customer interactions can happen in third-party apps across millions of mobile devices. Traditional data solutions are not designed for this dynamic new world, where the data can change continuously in volume, size and shape -- creating new requirements for accurate analytics.
Predictive Big Data Analytics Platform
Enabling enterprises to apply sophisticated analytics on fine-grained data, Apigee launched a major new version of its Insights big data analytics platform. The platform allows enterprises to deliver improved individualized services through predictive apps. “In today’s mobile-first digital world, it’s not enough to understand what your customers have done in the past– the most successful digital businesses will predict customer needs and take action to address them,” said Chet Kapoor, Apigee CEO. “Apigee Insights delivers predictive analytics at a level beyond that used at the Internet giants, and makes it available to any enterprise.”
The new Apigee Insights combines predictive analytics and advanced machine learning in a unified platform. It analyzes many types of data to build a comprehensive understanding of each customer, including consumer data, such as demographics and social media usage; product interaction data, such as how a customer interacts with a company’s products; and machine data that includes app and information services usage. It is available in industry solutions for verticals, custom solutions, and API solutions.
How Does it Work?
Apigee Insights stitches together data from a customer's API programs with data from internal systems and online data sources and delivers in-depth analysis and performance with a multi-channel data aggregator, distributed processing engine, intelligent data storage, analytic accelerators and expert services.
The platform can readily adapt to changing number, volume, size and sources of app economy data, and allows customers access to near real-time feedback loop to test, experiment and rollout changes immediately. Organizations can see insights throughout the entire app value chain or can choose to focus specifically on the context of the app user, the app developer, or on information analytics.
Think you have your app economy under control? Quickly tally up how many mobile apps and plug-ins you’re using to integrate with other systems to provide optimized customer experiences and then ask yourself, “do I know where all that data goes?” If you don’t, it might be time to think about how you might go about connecting with it all, nevermind figuring out what it all means.
To more clearly understand the Apigee Insights pattern, let’s break it down into its components:
The core flow:
The first thing to recognize is that Apigee Insights, unlike most explanations of big data, is about including big data and contextual insights into an interaction with customers, not into a traditional business intelligence pattern in which analysis is divorced from action.
The core flow Jhingran refers to is the interplay between a customer, one of many applications, and the back end systems of the enterprise. For example, in a web or mobile context, this core flow could be an interaction between a customer who has added a product to a shopping cart and is executing a transaction. The signals that the customer sends when they add a product to a shopping cart are quite strong. Much is known about this interaction.
The goal of the Apigee Insights platform is to support this core flow by adding contextual data captured from as many sources as possible, this is where big data comes in, and also from capturing data from the interaction that is intermediated by APIs.
The context:
There is a wealth of contextual data about customers that can come from a huge number of sources. This data is vastly different from the data that supports the core flow. Contextual data is noisy, incomplete, and contains only weak signals. The relevance of many of these signals evaporates quickly. Contextual data can come from any number of sources. It can tell you what other sites a customer has visited. What comments they have made on social media. How often they have interacted with your company or your partners. These signals are found in huge repositories of data that must be mined.
The Apigee Insights pattern suggests that this data is best used at the time of interaction. These signals should be available in distilled form for use by applications, not hidden away in a data warehouse.
The interaction:
As the world of mobile apps grows more important as a channel, and the interactions with partners must be supported in sophisticated ways, APIs are becoming the layer that presents a company’s capabilities to the outside world. While most organizations are thrilled with the flood of creativity and innovation that can be enabled by an API, too often, the fact that APIs can be a important way to learn about customers and partners is ignored. By tracking the interaction between a customer through APIs and by making the applications smarter with contextual data, you have an opportunity to operationalize insights. In other words, you can use many of the weak signals before their value has evaporated.
The analysis:
Two types of analysis are involved in the Apigee Insights pattern. The first is the gathering of weak signals from big data where ever it is found. The second is the real time decision making inside an application to use both strong and weak signals to best respond to customers. Both types of analysis are crucial to operationalizing big data.
The action:
Based on the signals inside Apigee Insights, the the APIs and the applications now have a wide collection of signals to work with. They can use these signals to provide better service to customers.
With this background, we can know understand what Apigee Insights provides, which is an intelligent layer of APIs, data, and analytics at the edge of the enterprise. The chatty interaction between customers can be handled with all the contextual data to make the interaction smarter. The context can be enhanced by the signals sent by customers through use of APIs.
Apigee Insights is built to support the next generation of applications that will combine both the power of the strong signals from internal systems and data with the weak signals from external data, contextual data, and interactions. The architectural principle of Apigee Insights is that this sort of intelligence cannot live on the back end or the front end, but must live in the middle, at the edge.
One of the least satisfying aspects of most discussions of big data is a lack of attention to patterns of value creation. Most of the time, we are presented with scenarios in which data is poured into Hadoop or sifted with Splunk and somehow you get insights.