Providing employees with charge cards and expense accounts is fine when it comes to empowering employees to do their jobs effectively. However, it can create challenges for any company that wants to understand how and where its money is being spent. Employee Generated Spend (EGS) needs to be captured and reconciled with other spending channels within the organization, such as procurement, accounts payable, and travel management.
Moreover, EGS may involve a wide variety of expenses – from the acquisition of ergonomic chairs and office equipment for a home office during COVID-19 lock-down to meal delivery, entertainment, and more. Employers want some level of visibility into those expenses – not just to control costs but to ensure good governance and accountability to shareholders. Using data to identify hidden costs and efficiency opportunities, PredictX offers solutions to provide visibility into how to meet these challenges. Furthermore, they can assist companies in finding better ways to optimize travel-related expenses and incorporate these trends into future strategies.
PredictX built its EGS solution using proprietary Artificial Intelligence (AI), advanced data analytics tools, and Vector – a high-performance columnar database. The solution combines and analyzes data from a wide range of systems and sources – including travel, expense cards, HR, and meeting data systems – to provide managers with opportunities to control costs and optimize employee spending programs.
While PredictX had a cloud-based offering on its long-term road map, an opportunity presented with a large client that had a specific requirement for the ability to deploy on the Google Cloud Platform (GCP). The current solution delivered all the business value this customer wanted, but GCP infrastructure provided necessary security and governance.
What could have been viewed as an insurmountable barrier, executives at PredictX recognized an opportunity. PredictX had invested in Vector database for its analytics processing due to Vector’s unbeatable performance capabilities. For this reason, the team needed a way to provide that same unmatched performance in the cloud. This led the team to begin investigating cloud alternatives, including the Avalanche platform.
Migrating their existing Vector database to the Avalanche Cloud Data Platform would allow PredictX to preserve existing business logic and code while still proving a cloud-based solution to customers.
The Avalanche Cloud Data Platform is a fully managed data platform that provides integration and warehouse services to deliver high performance and scale across all dimensions – data volume, user concurrency, and query complexity – at a fraction of the cost of other solutions. It is a true cloud data platform that can be deployed on-premises as well public clouds, including Amazon Web Services, Azure, and Google Cloud Platform. Supporting hybrid deployments enables organizations to migrate or offload applications and data to the cloud at their own pace.
For PredictX, this allowed them to easily move their existing on-premises Vector connectors to the Avalanche platform – a task that would have taken a great deal of time spent re-writing and re-coding operations.
Connecting the Avalanche platform to the tracking systems where the new client’s EGS data existed was very simple and straightforward. Similarly, the AI and predictive analytics tools that PredictX had been running on Vector also easily connected to the Avalanche platform without needing any major modifications. Best of all, there was no impact to performance. In fact, PredictX showed a performance improvement running its platform in the cloud.
The PredictX platform relies heavily on metadata, workflows, and orchestration routines originally designed for on-premises deployment. PredictX developers had some concern about the master data and workflow routines forking in different directions if they needed to make major modifications in order for them to run on the Avalanche platform. Actian professional services helped PredictX use the same master data and workflow routines, which would enable greater deployment flexibility and provided a cloud-based option for PredictX customers going forward.
Given that the PredictX platform had not been deployed in the cloud before, they expected that it would be several months before PredictX would be able to provide access to their PredictX platform on the GCP cloud. Initially, PredictX anticipated that it would be 6 or 7 months before the client would be able to conduct any user acceptance tests. The migration and configuration was straightforward and their customer was able to access the UAT environment in just 3 months.
“We had multiple workflows and management routines that were built over years to run Vector on-premises,” says Andrew White, Executive Vice President of Solution Engineering at PredictX. “We had been a bit concerned about how long it would take to get them up and running on the Avalanche platform, but they pretty much worked right out of the box. The migration was completely invisible to our end users, and that was exactly what we hoped.”