Call centers are embracing automation. There’s debate as to whether that’s a very good thing, nevertheless it’s happening — and quite possibly accelerating.
In line with research firm TechSci Research, the worldwide marketplace for contact center AI could grow to just about $3 billion in 2028, from $2.4 billion in 2022. Meanwhile, a recent survey found that around half of contact centers plan to adopt some type of AI in the following 12 months.
The motivation is slightly obvious: Call centers want to reduce costs while scaling up their operations.
“Corporations with heavy call center operations, seeking to scale quickly without the constraints of human contact center agents, are highly receptive to adopting effective AI voice agent solutions,” entrepreneur Evie Wang told TechCrunch. “This approach not only reduces their overall costs but additionally decreases wait times.”
Wang is certainly one of the co-founders of Retell AI, which provides a platform firms can use to create AI-powered “voice agents” that answer customer phone calls and perform basic tasks corresponding to scheduling appointments. Retell’s agents are powered by a mixture of huge language models (LLMs) fine-tuned for customer support use cases and a speech model that provides voice to text generated by the LLMs.
Retell’s customers include some contact center operators but additionally small- and medium-sized businesses that recurrently take care of high call volumes, like telehealth company Ro. They will construct voice agents using the platform’s low-code tooling, or they will upload a custom LLM (e.g. an open model like Meta’s Llama 3) to further tailor the experience.
“We invest so much within the voice conversation experience, as we see that as probably the most critical aspect of the AI voice agent experience,” Wang said. “We don’t view AI voice agents as mere toys that one can create with a couple of lines of prompts, but slightly as tools that may offer substantial value to businesses and replace complex workflows.”
Retell worked well enough in my transient testing, not less than on the call-facing side.
I arranged a call with a Retell bot using the demo form on Retell’s website. The bot walked me through the technique of scheduling a hypothetical dentist’s appointment, asking questions like my preferred date and time, phone number and so forth.
I can’t say the bot’s synthetic voice was the most effective I’ve heard when it comes to realism — actually not on par with Eleven Labs or OpenAI’s text-to-speech API. (Update: Wang tells me Retell’s using a custom ElevenLabs voice, which could explain the lower quality.) Wang, in Retell’s defense, said that the team’s been mostly focused on reducing latency and handling edge cases, like interruptions which may occur in a conversation.
The latency is low: In my test, the bot responded just about without hesitation to my answers and follow-up questions. And it stuck to its script. Try as I would, I couldn’t confuse it or prompt it to behave in a way it shouldn’t. (After I asked the bot about my dental records, it insisted that I speak with the office manager.)
So are platforms like Retell the long run of call centers?
Possibly. For basic tasks like appointment scheduling, automation makes lots of sense, which might be why each startups and large tech firms alike offer solutions that compete head-on with Retell’s. (See Parloa, PolyAI, Google Cloud’s Contact Center AI, etc.)
It’s low-hanging — and seemingly revenue-generating — fruit. Retell claims to have tons of of shoppers, all of that are paying per minute of voice agent conversation. Retell has raised a complete of $4.53 million in capital up to now, courtesy of backers including Y Combinator (where the corporate was incubated).
However the jury’s out on more-complicated queries, particularly given LLMs’ tendency to make up facts and go off the rails even with safeguards in place.
As Retell’s ambitions grow, I’m curious to see how the corporate navigates the various well-established technical challenges within the space. Wang, not less than, seems confident in Retell’s approach.
“With the arrival of LLMs and up to date breakthroughs in speech synthesis, conversational AI is getting ok to create really exciting use cases,” Wang said. “For instance, with sub-one-second latency and the power to interrupt the AI, we’ve observed users speaking in fuller sentences and conversing as they’d with one other person. We’re attempting to make it easy for developers to construct, test, deploy and monitor AI voice agents, ultimately to assist them achieve production readiness.”
