Observe.ai, a venture-backed developer of contact center software, today introduced a large language model optimized to streamline customer service tasks.

The startup refers to the model as the Contact Center LLM. It powers a new set of automation features that Observe.ai is rolling out in conjunction under the name Generative AI Suite.

Observe.ai, officially Z21 Labs Inc., is backed by more than $210 million in venture funding. It provides a platform that helps make the contact center teams more productive. According to the startup, its installed base includes more than 200 organizations. 

Observe.ai’s platform uses artificial intelligence to analyze customer interactions and highlight areas for improvement. It can, for example, point out if help desk representatives could present users with upselling offers more often. Observe.ai also promises to ease a number of related tasks. 

The company’s new Contact Center LLM is available in three versions with 7 billion to 30 billion parameters, or configuration settings. Parameters determine how an AI processes data. Often, the more such settings are implemented in a neural network, the more accurate it becomes. 

Observe.ai says that it trained the Contact Center LLM on a dataset that includes hundreds of millions of customer interactions. According to the startup, the model is more accurate than many general-purpose neural networks that were created using less specialized data.

In an internal test, Observe.ai compared Contact Center LLM with GPT-3.5, one of OpenAI LP’s most advanced general-purpose AI systems. The company claims its model can perform customer sentiment analysis up to 35% more accurately. Moreover, it says, enterprises may train the Contact Center LLM on their own custom datasets to further improve accuracy.

The model powers a new set of features called Generative AI Suite that Observe.ai introduced for its platform today. The features promise to ease several time-consuming aspects of contact enter teams’ work.

When a customer asks a complicated question, help desk teams must often consult their company’s internal knowledge base to find the answer. Manually sifting through knowledge base articles for information can take a significant amount of time. According to Observe.ai, its new Generative AI Suite automatically surfaces answers to customer questions.

The suite also automates the task of producing support call summaries. It can organize a summary as a series of bullet points or in the form of a single paragraph. The text is available in the Observe.ai interface, as well as via an application programming interface that makes it possible to stream the data to external analytics tools. 

“Agents can simply type the question and get ready-to-use answers,” Satarupa Chatterjee, Observe.ai’s senior director of product marketing, explained in a blog post. “Agents also get citations in the form of links to the original documents if they want to access more details around the answer.”

Observe.ai says its new language model eases employee training as well. According to the company, the Contact Center LLM can analyze each support agent’s customer interactions and automatically generate customized coaching notes. Observe.ai says its technology allows companies to provide contact center teams with feedback more quickly than would otherwise be possible. 

Image: Observe.ai