USA News Hub

Reltio is getting into the Customer Data Platform (CDP) space today with a new release that extends the company’s master data management (MDM) platform.

The Reltio Customer 360 Data Product is built on top of the Reltio Connected Data Platform which has been increasingly adding enhanced artificial intelligence (AI) capabilities to help improve MDM. While MDM is about mastering and having the most accurate version of all types of data, the new Reltio Customer 360 Data Product is specifically focussed on providing organizations with a unified view of customers.

The product is adding a series of enhanced AI features, including the Reltio Intelligence Assistant (RIA) which is a generative AI conversational assistant for data exploration. Reltio is also using a large language model (LLM) powered approach that it has developed called Flexible Entity Resolution Networks (FERN) for ensuring data record accuracy.

“What we have seen is a greater need for a 360 degree view of customers as a horizontal need across every industry,” Manish Sood, CEO and Founder of Reltio told VentureBeat in an exclusive interview. “They’re all struggling to stitch together the holistic, unified 360 degree view that they not only need to use for marketing, which was the original customer data platform type of promise, but they need that as an institutional interoperable asset that can be brought up in a real time orientation.”

How FERN improves entity resolution for mastering customer data

A critical aspect of mastering data is being able to understand all the different ways a specific entity is present in a data set.

For example, a customer’s name might be spelled in a number of different ways (e.g. Dave or David). Entity resolution is the operation that helps to figure it all out.

The Reltio Customer 360 Data Product integrates the company’s FERN approach which uses the power of LLMs to help accelerate entity resolution. Sood explained that FERN uses LLMs to enable zero-shot entity resolution capabilities. This removes the need for any training of models for each customer context.

FERN uses pre-trained open source large language models under the hood to do fast and highly accurate string comparisons for entity matching.This allows it to do cross-language matching of entities like names, addresses, etc. that may be represented differently in various languages or even character sets. He noted that it can perform entity resolution in under 50 milliseconds, making it suitable for real-time use cases.

Sood explained that Reltio is using a number of different open source models to enable FERN, though he did not identify the specific models.

“What we have done behind the scenes is strung together a few different models that can be used for some very specific functions like name, address, company  or product name comparisons. and broken that down by different models,” he said.

Don’t know SQL? that’s ok RIA can help

Reltio is also using multiple models to help enable its Reltio Intelligence Assistant (RIA) technology that is part of the Customer 360 Data Product.

Doing advanced data queries on various types of data platforms in the past has typically involved using structured query language (SQL). RIA removes that requirement providing a gen AI assistant that enables natural language queries of data.

“Our goal is to make it easier and simpler for even the non-technical users to be able to come in and interact with the data and explore various facets of it easily,” Sood said.

Reltio looks beyond regular CDP features

With a regular CDP, data is typically used for customer marketing and management. Reltio is positioning its new offering as going beyond typical CDP capabilities as a data product.

The concept of a data product is all about taking data and using it as a product to solve a specific purpose. Sood explained that from Reltio’s perspective a data product takes specific domain data and packages it for easy, consistent consumption across different use cases and downstream systems. That packaging includes capabilities for real-time API access, integration into analytics systems as well as providing AI models with up-to-date enterprise context.

Going a step further Reltio is providing data observability as part of the Customer 360 Data Product. That includes constant monitoring of data as it is being served by an API as well as maintaining service level objectives and consumption patterns.

Looking ahead, Reltio plans to expand its offerings to include additional data domains by vertical and use case. However, the current focus remains on refining the conversational experience based on customer feedback.

“I’m really excited about spending more time with the customers and how we should further refine the experience and what kind of additional skills or features we should add so that they are able to explore data in many different ways,” Sood said.