In an era where profit margins are thinning, banks and asset managers are looking to derive more value from their data to provide better returns, and they need to generate signals quickly in this competitive market. It is for this reason that Kx, a division of First Derivatives, stood out at the AFTAs, winning the infrastructure section of the most innovative third-party technology vendor category, thanks to its suite of services. Kx technology incorporates the kdb+ time-series database, which provides improved processing performance, in-memory computing, streaming analytics and operational intelligence.
James Corcoran, senior vice president of products, solutions and innovation at Kx, says the vendor recently helped clients build systems for trade surveillance, regulatory compliance and transaction reporting. Its platform is designed to assist with challenges associated with capturing, processing and storing huge volumes of real-time and historical data.
In the past year, Kx has continued to invest in making kdb+ faster and has added new features, including machine learning capabilities and Python integration. This has been done through kdb+ 3.6, the latest version of its database and analytics application.
Previously, kdb+ worked best with structured and time-series data, but the new version extends its capabilities to analyze text and different types of unconventional data. This allows developers to combine structured and unstructured data within the kdb+ database and analyze them both at the speed kdb+ is known for.
The integration with Python allows programmers working on machine learning applications to use kdb+, even without prior knowledge of the database. Similarly, kdb+ programmers can integrate Python into their applications. “We’ve also released our own native machine learning, feature engineering, and natural-language processing libraries, which opens up our platform to a wide variety of use-cases such as analyzing email, instant messaging and chat-room data for the purposes of communications surveillance,” says Corcoran.
Kdb+ is also available on-demand, which is useful for clients wanting to run burst computing workloads, says Corcoran. This enables running any scale system on a cost-effective and elastic per-core-minute basis. Corcoran sees an increased demand for time-series database solutions as organizations now rely more on real-time data to inform, and in some cases automate, decision making. “We are also seeing increased cloud adoption alongside a move to develop applications using predictive analytics and artificial intelligence from the ground up,” he says.
Looking to the future, Kx plans to leverage its new capabilities and extend that across its broader solution set.
Bryan Cross, who heads UBS Asset Management's QED group, joins to discuss alternative data and AI.Subscribe to Weekly Wrap emails