Ad, a Toronto-based customer service automation startup, has long been using big linguistic models in its solutions, but today the company is announcing a new suite of tools powered by generative AI that aims to take that automation to the next level. to another level.
The company’s founder and CEO, Mike Murchison, Generative AI hopes to help its customers get closer to that ideal.
“In short, in ada, build once and through any channel message, voice, [whatever]And that bot is spread everywhere. The initial build time for the bot is essentially zero because we’re creating an answer based on the expertise within your company,” Murchison told TechCrunch.
It’s a tall order, especially with the problem of hallucinations in large language models that don’t know what to make of words or actions from cognitive data.
“One of the most difficult problems to solve in implementing generative AI at work right now is the security problem. How do you ensure that?” [the bot is] Is it safe, accurate, and important to give the right answer? How do you actually do that? ” he asked.
He said these issues are not part of the LML problem. “While there’s a lot of complexity, we’ve built a lot of IP around our overall AI pipeline [give an answer] With confidence. That’s a big, big reason why our customers are so excited about it,” he said.
What’s more, he says the amount of time you need to delegate complex tasks to a human falls significantly with this product, but says Ada will integrate with customer service software like Zendesk and Salesforce when necessary.
“In that case, where there is a gap, we will provide it to the person without problems. And that was a big part of Ada’s growth, really; Especially in the last five years [we have seen] “How effectively we’ve filled agent desktops in this space, especially Zendesk and Salesforce,” explained Murchison.
Today, the company has more than 300 customers using the platform, including Meta, Verizon and Shopify. In the year Launched in 2016, the company has raised more than $190 million, including a $130 million Series C in 2021.