As we begin the new year with our lists of professional resolutions and goals, it can be interesting and beneficial to hear from industry leaders about the challenges that they are facing and their approach to successfully address them. In December, at our Insight-Driven Organization Event in London, many insights, derived from hands-on implementations, on how to become an insight-driven organization were presented and discussed by esteemed representatives from Intel, UBS, Gresham, and ActiveViam.
The concept of evolution vs revolution and the recommended optimal path towards gaining instant insights with an existing big data ecosystem in a simple manner resonated well with the audience. Questions asked by leaders from multiple industries such as financial, transportation, retail, insurance and more included “how can my organization leverage machine learning?” and “how do I select a technology stack to build an analytics ecosystem?” The importance of data consistency, simplicity and TCO were stressed as important prerequisites for big-data and advanced analytics projects.
We’ve compiled a recap of the highlights here for you who could not attend.
Jason Turner from Gresham presented a solution for building resilient, highly available cost-effective systems for heavy data loads especially when there are data spikes, which can be a challenge for high speed, in-memory computing based solutions in industries that demand rapid response times and “always-on” service. “Leveraging a multi-tiered data storage model where the ‘hot data’ is stored on RAM and historical data is stored on an external SSD or flash array has addressed the increasing data volumes and data spikes in a cost-effective manner that meets the demanding speed and service levels that our customers require.”
Technology innovations from Intel support this architecture and Justin Wheeler from Intel addressed this ever-growing need for “Big Memory and Fast Data,” and the quest for the “ideal memory”—big, affordable, fast, and persistent memory that would allow for real-time access and advanced analytics on large databases. Intel described how the insight-driven organization requires more memory for machine learning and AI workloads; Optane technology extends the working set by as much as 8x, and more cost-effectively than what you would pay for DRAM.
Consistent data from multiple sources serving various applications is the base of many exciting services, but it can be challenging to design and deploy. Adam Laczynski from UBS spoke about the requirement for scale and data consistency and referenced how the UBS TESS project, a distributed scalable post-trade processing platform running on 300 JVMs supporting tens of bank applications which share the same data was designed and deployed. Tens of millions of events per day are processed, to reconcile millions of trades per day and to assess risk for millions of exceptions per day.
“TESS achieved superior real-time performance with 100% uptime over the last few years and is able to handle peak loads at events such as the 2016 US elections and Brexit.”
Fabien Celier from ActiveViam once again “energized” the audience as he did in our Paris event in October where he demonstrated how organizations are discovering data insights via multi-dimensional analysis and visualization leveraging the integration of Excel Table Pivot, ActiveViam’s ActiveUI and GigaSpaces InsightEdge. The demo showed the many valuable insights and direct impact derived from the use of electricity patterns in the different geographical areas and is just an example of how many industries can gain insights from their data today.
Tal Doron from GigaSpaces spoke about how companies can gain a competitive edge with In-Memory Insight platforms through fast data, real-time analytics, and mission critical transaction processing. “The value of the data decreases as the gap grows between the moment the data arrives and the point of action taken in response; from a business point of view, we need to analyze this data very fast and as close as possible to the source,” said Tal. “Investments in AI innovations, specifically deep learning are growing rapidly across industries such as financial services, healthcare, transportation, and retail to reap insights that can predict and impact directly on business workflows and actions.” Tal presented a live demo of GigaSpaces InsightEdge utilizing Intel’s BigDL framework.
The combination of thought leadership, technical content, and demonstrations resonated with the audience of enterprise executives. We look forward to seeing you at one of our upcoming executive events across the globe.