Copyright © Emerce 2023
Existing data structures like CRMs and CDPs are no longer sufficient in an age where AI is foundation, lubricant and enricher of customer data. CM.com developed a platform for this new reality.
During the Human After All event, the Breda-based company launched a new data platform at a customer day that helps organisations organise their data in such a way that AI can deliver more value in customer contact and personalised communication, among other things.
The essence of the new platform is, that the marketing world is no longer a sum of separate instructions, but becomes a whole in which generative AI understands context and performs tasks.
This sounds abstract, but ceo Jeroen van Glabbeek (photo) explains it conceptually simply to Emerce.
"It's about context mining. Take Prénatal as an example. The 'customer', the person concerned, has not yet been born. But so you do know that there are parents and that those are people with tastes that suit them. You bring this kind of data together in a way that support and marketing can draw from. They also enrich the customer view back and forth."
Data platforms like CDPs and CRMs can only work with data that has been captured. That the systems know. The new data platform can make inferences: draw conclusions about unique situations where not all the facts and data are known. In other words, the computer colours the picture in the most logical way. Context mining.
"But you can also think of sports associations. An AI can distinguish between the parent paying the bills, the sporting child, the association and the club where someone is a member or volunteer."
"The new platform brings this context together and makes it readily available to AI agents. With AI as a new user of CRM, ERP and CDP data, fragmentation no longer works. That's why we developed a hybrid data platform that brings all relevant information together in one structure. AI works with the full context."
Context is the new magic word in the GenAI world. Context is crucial for several reasons. Without it, an AI model cannot determine what you are really asking. For example, the phrase "It's too big" without context could be about clothes, a house or an apple pie. In addition, AIs' memories are getting bigger and bigger. If the machine has no meaningful context, in longer conversations a model may forget what the conversation topic actually was. And good context is the basis for relevant advice, nuanced answers and questions.
Three customers of CM.com currently have access to the new system, which is in its alpha phase. Van Glabbeek says it will be launched as a finished product in the summer.
After the summer, he expects to launch a product that can predict targeted customer queries. The basis for this lies in a small, task-oriented AI model with a built-in context understanding engine.
© The Content Exchange, source News