Business intelligence platform GoodData is preparing for its next wave of rapid growth, releasing new capabilities and SDKs for both customers and third-party "white label" developers. The new releases coincide with new funding and executive hires, all of which suggest GoodData is a key dashboard and data analytics product to watch heading into 2015. ProgrammableWeb spoke with Cory Vander Jagt, director of product marketing at GoodData, about the new features.
GoodData is a business intelligence platform that allows end customers to collate, store, analyze and visualize a wide variety of private and SaaS-based data sources. The platform also works with partners and developers to provide a “Powered by GoodData” service whereby developers create apps and visualizations using GoodData’s data analytics engine and visualization capabilities.
Following recent presentations at API World and DataWeek in San Francisco, GoodData has launched an update of its Extensible Analytics Engine (XAE) and new SDKs for automation and visualization.
The new features help GoodData position itself in a market that expects greater predictive features and big data analytics capabilities that can update and rerun queries in real time, without requiring manual rewriting of the underlying data models.
“First and foremost, people use us for analytics and for presentation of any data from any source,” says Vander Jagt.
For our Extensible Analytics Engine, we have updated with adaptive intelligence features, so that when customers update their data models, we dynamically create the optimized SQL on the back end. That makes cost-efficient use of their queries and makes use of our multitenant platform: it's running in memory, so it will make most efficient use of the path and return results quickly.
If you want to change your query, you can filter by any attribute: by source channel, product, geography, maybe you want to look at something coming from your e-commerce channel rather than from your social media. Then, if you want to update your visualizations, users with the correct permissions can go in on ad hoc basis and see those investigations, make those calculations on the fly and see the results instantaneously.
The ability to not only do those ad hoc changes and adjustments to visualization, but say you want to go after a new channel, now we are starting to do paid advertising on Pinterest, I can add that to the model now. XAE can programmatically add that without having to update each model, each visualization screen, and that’s what saves our customers a heap of operational time.
Adding More Predictive/Forecasting Tools
Like other analytics services, GoodData needs to look at how to leverage the emerging technologies of machine learning and predictive analytics into the services it provides its customers. In the past, they have been able to help conduct hypothesis testing, a form of predictive analytics based on historical data in order to determine possible future forecasting outcomes of reapplying strategies that have proven impactful.
Vander Jagt sees several ways that GoodData is stepping up to this challenge:
One of the features in the new XAE we are most excited about is the advanced statistical functions. This includes trend-line functions that let you do some forecasting.
One of our core capabilities is that with any of the information we are pulling in, we are very good at snapshotting: We can provide a snapshot of a point in time for every historical data load. As we continue to do that, the better calculations we can make, and the more accurate predictions that we can forecast.
We also have full end-to-end data integration, which means that as we process the information and take unstructured information from social media, we can perform sentiment analysis internally, and we also allow users to integrate and run it through any other third-party predictive analytics APIs. So you can use third-party machine learning tools in GoodData.
One of the key reasons we are chosen to work with large brands is because we can combine structured and offline data and do advanced statistical analysis like projections. We can then set proactive notifications on any core metric. So, for example, customers can set any threshold you want to manage, for example, setting an alert if a negative sentiment analysis level is reached in any campaign. The idea is that our analytics are leading you through the phases from informative to actionable, proactive decision-making.
Along with the more advanced query and analytics capabilities, GoodData has released two new SDKs. The Automation SDK allows easier integration of the new XAE features of multithread processing and automatic updating of big data search queries.
Vander Jagt points to one use case of the new Automation SDK, where a user was able to perform and update analysis on 5,000 Zendesk projects in around 10 minutes. It is a process that would normally have taken 10 hours with the API, or 28 hours manually, GoodData estimates.
Vander Jagt explains:
Our SDKs make it much easier to manage the integration and end-to-end management.
GoodData’s focus on building out its open analytics platform has drawn the attention of investors and key data analytics influencers. Today, GoodData has announced an additional $25.7 million in series E funding and added Pete Godbole (ex-VMware) and Philip Carty (ex-MongoDB) to its executive team.
The moves come as GoodData is increasingly being used by enterprises, including Hewlett-Packard, Nordstrom and Virgin America.
Developers can trial the platform for free or investigate becoming a third-party developer partner and creating products on the platform that utilize the XAE and new SDKs.