Io-Tahoe LLC, a unit of British utility giant Centrica PLC, today is enhancing its machine learning-driven data catalog with features that it says can discover and classify streaming data.

The capability can be used, among other things, to discover personally identifiable data while it’s still in motion to enable better regulatory compliance.

The privately held company, which was founded as Rockitt Inc. in 2014 and sold to Centrica three years later, has built a platform that it says can find and classify data across a wide range of platforms ranging from traditional databases to semi-structured data lakes. The software can also identify data in some unstructured formats, but the company didn’t provide details. Most catalogs require human intervention to apply metadata – which is data about data — to underlying information, but Io-Tahoe says it has automated this process with patent-pending technology.

The capabilities in the new Smart Streaming Discovery catalog are well-timed for the regulatory environment that has been growing increasingly stringent since the General Data Protection Regulation went into effect in Europe nearly a year ago. Several countries and U.S. states have said they are considering similar regulations. The California Consumer Privacy Act, which is scheduled to go into effect next January, has been called “America’s GDPR,” with stringent regulations on how data is collected and used.

One of the problems organizations face in managing customer data is that they don’t always know where it is or what it looks like. Variations in format and labeling, as well as fragmentation of data across a large organization, make it difficult for an organization to catalog all of its information assets.

With Io-Tahoe’s new capabilities, said Rohit Mahajan, the company’s chief technology and product officer, “not only will organizations know what data they have and where it is located, they will now have the ability to understand what data is sensitive and flag it before it lands in data stores.”

The current iteration of the company’s streaming discovery product uses Apache Kafka, but Mahajan said it will work with any streaming analytics platform. “We have developed a proprietary technology that can take data at rest or streaming – it doesn’t matter to us – and tag it and call it out,” he said. The platform comes out of the box with the ability to flag data designated as sensitive under GDPR rules and it can be modified by users for their own purposes, Mahajan added.

Propelled by the internet of things and scalable streaming platforms such as Apache Kafka, the market for streaming data is expected to grow rapidly. Research and Markets Ltd. expects the streaming analytics market to grow more than 32 percent annually to reach $21.6 billion by 2023.

Io-Tahoe’s technology uses “various hashing and complex bit-mapping techniques” to flag suspect data and holds it in cached storage pending the application of rules about what to do with it, Majahan said. “Once it’s persisted onto customer data stores, they may want to make it part of mainstream processing, but that’s their decision,” he said.

The combination of caching and high-speed rules-based analysis prevents data loss during the process, he added. Data flagged as unacceptable for an organization to store under local regulations can be marked for automated disposal.

The company is positioning its new product as both a compliance and a data loss prevention tool. The latter is useful in detecting data breaches, in which personal information is often siphoned away by attackers. The Io-Tahoe platform is said to use deep-learning techniques to automatically find and tag sensitive data during transmission to enable faster remediation.

Pricing information wasn’t released.

Image: Pixabay