Five9 has launched an “industry-first” GenAI Studio that allows users to optimize the GenAI-powered capabilities they leverage within their unique contact center environments.
To understand how, it’s first critical to know how every out-of-the-box GenAI application works.
Essentially, a pre-configured prompt feeds a large language model (LLM). The LLM then generates a response or “output”.
Often, these default prompts work well. But, as use cases become more complex and ask more of the LLM, the accuracy of the outputs drops significantly – and that’s where the GenAI Studio comes into play.
It offers a “prompt playground” that allows users to create prompts for a specific use case, evaluate the performance of the prompt using in-house data, and optimize it.
The contact center may also use the playground to test different LLMs and isolate the model that performs best for a particular GenAI use case within their environment.
To start, Five9 will only make one use case available within the GenAI Studio: “Agent-Assist Summaries”.
That use case will come with suggested prompts for a post-contact summary that captures the customer’s intent and sentiment. The prompts may also standardize the format of the summary.
Yet, contact centers can go into the GenAI Studio to tweak and test the prompts and various LLMs – gauging their performance in summarizing historical contact transcripts.
After running these tests, the contact center can optimize the feature’s performance and draw maximum value from the use case.
According to Five9, all this takes is “just a few clicks” – thanks to a low-/no-code interface that allows customer experience teams to use the GenAI Studio – lowering the burden on developers.
Over time, the Studio will power all Five9 GenAI use cases. Indeed, Agent-Assist Summaries is just the beginning.
Soon, Five9 will create a deep library of use cases and sample prompts. Customers may then continually trial, enhance, and embed GenAI features into their contact centers without the need for prompt engineering experts.
Excited to share this vision, Jonathan Rosenberg, CTO and Head of AI at Five9, stated: “With GenAI Studio, customers can infuse off-the-shelf models with their unique data, making the Generative AI model their own.
Delivering this type of capability with both power and ease has been a challenge until now; it’s in line with the core value proposition that Five9 has delivered and will continue to deliver in the future.
“GenAI Studio is the next big step in our journey to transform the contact center for the future.”
The Studio is also a step forward for Five9 in delivering responsible AI, as it can provide clients with closer control and monitoring of GenAI outputs.
Moreover, by offering an LLM-agnostic layer, the CCaaS stalwart doesn’t constrain its customers to a single model. Instead, they may select the option that aligns best with their business needs, whether that’s a proprietary model – like ChatGPT or Gemini – or an open-source model – such as Llama
Such innovation within the GenAI Studio enthused Dan Miller, Founder & Lead Analyst at Opus Research.
“I’ve been impressed by Five9’s strategy to anticipate and meet demand for AI in the contact center through a combination of acquisitions and internal development,” he said.
They took an early leadership position with market-ready AI solutions, and, with GenAI Studio, they provide easy-to-use tools that leverage their formidable array of resources to build, train, and test highly personalized virtual assistants.
Finally, the GenAI Studio will launch in the second half of the year, perhaps at its annual CX Summit.
There, attendees may also wish to dive deeper into Five9’s 2023 acquisition of Aceyus, which may have proven the springboard for Five9 2.0.
