In our previous post, How In-Memory Computing is Accelerating Business Performance, we explained the disruptive potential of in-memory computing. Performance gains resulting from faster execution of queries were one of the top benefits mentioned. However, in-memory computing goes way beyond performance gains, allowing organizations to do things differently and achieve new levels of competitiveness. This post illustrates this with a few examples.
For the last 30 years, IT teams have tried to balance the growing need for interactivity, while finding workarounds to offset the slow performance of disk-based databases. Typically, hours if not days pass from the moment data is generated by transactional applications to when it’s available for analytics. In-Memory computing shrinks this time laps, eliminating the border between transactional and analytical processing and allowing organizations to invent new and improved business practices.
Dynamic pricing has become a differentiation weapon in the ecommerce world, where consumers have the power to choose when, where, and at which price they will purchase goods and services. To remain competitive, internet retailers must quickly revisit their prices to respond to supply and demand variations, market changes or competitive positioning. To fulfill this requirement, dedicated pricing teams constantly analyze pricing performance and provide recommendations on price adjustments.
In ecommerce, however, time is of the essence. Price adjustments become meaningless if they are based on outdated and incomplete data. Reality is that price specialists sometimes spend two days collecting data from operational systems (competitors prices, inventory data, web traffic information) before they can provide any recommendation.
Yet with in-memory analytics price specialists can run their analysis on live data as it is created and updated by operational systems. They can immediately respond to external events, as they happen. When a competitor drops the price of a star product, the event is immediately detected and responded to with the appropriate price adjustment, after taking into account factors such as inventory levels, and commercial strategy (price leadership, margin increase, etc.). The result is an extremely agile pricing with a direct impact on revenues, margins and market share. Without in-memory analytics, the discipline of dynamic pricing could simply not materialize.
Management by exception is based on the usage of metrics and indicators to focus attention to any deviation from normal business operations. This practice is sought after by logistics firms, whose supply chains are growing in complexity, forcing them to focus on the most critical events in order to sustain operational performance and customer satisfaction.
Consider a large third-party Logistics Service Provider that transports one million vehicles from factories to a network of dealers around the globe. Deliveries must adhere to strict deadlines to stay within SLA ranges and avoid penalties. In turn, supply chain managers must track multitude of unqualified alerts (delays, works, faults, bad weather conditions). While some of these events get resolved, others require their full attention. In that case, supply chain managers do not want to miss that one incident which may appear as trivial at first sight but which could potentially threaten the delivery of a large order. In short, they want to be able to determine the events that deserve their attention.
With In-memory analytics (combined with CEP) in place critical events are handled seamlessly. Key performance indicators are computed on the fly as soon as an update takes place in the underlying operational systems, and managers are notified of any meaningful deviation. Live data is combined from multiple sources “traffic, vehicles, destinations, SLAs“ allowing managers to instantly assess the impact of an incident on other downstream operations. By providing a mechanism to score alerts and eliminate noise, supply chain managers can focus their attention on the most meaningful events. This in turn reduces operational risk and improves customer satisfaction.
Dynamic pricing and supply chain management by exception are only two examples illustrating how in memory analytics fosters real-time decision-making and enhances agility and competitiveness. For additional examples, see our customer case studies.
I am Head of Research and Development at ActiveViam. I joined ActiveViam soon after its creation back in 2005 and have been leading the team in charge of designing, developing and supporting ActiveViam’s in-memory analytics solutions. As one of the first employees, I was heavily involved in the design of ActivePivot, ActiveViam’s in-memory OLAP engine.
Before joining ActiveViam, I worked several years for a consulting firm specialized in the financial sector. From my years in consulting, I developed a strong customer orientation and I am keen on keeping a close eye on customers’ use cases. I graduated from Ecole Polytechnique and Telecom Paris.
I am Head of Research and Development at ActiveViam. I joined ActiveViam soon after its creation back in 2005 and have been leading the team in charge of designing, developing and supporting ActiveViam’s in-memory analytics solutions. As one of the first employees, I was heavily involved in the design of ActivePivot, ActiveViam’s in-memory OLAP engine. Before joining ActiveViam, I worked several years for a consulting firm specialized in the financial sector. From my years in consulting, I developed a strong customer orientation and I am keen on keeping a close eye on customers’ use cases. I graduated from Ecole Polytechnique and Telecom Paris.
Innovative data analytics that empower financial services firms to be ready for the unexpected.