This article first appeared in Data Sheet, Fortune’s daily newsletter on the top tech news. Sign up here.
I was in Atlanta Thursday, and for the second time in two weeks I was reminded that Silicon Valley has no monopoly on innovation.
I visited a company adjacent to Georgia Institute of Technology called Pindrop, which makes voice authentication and security products used by financial services companies and the like to cut down on fraud. Its AI-driven software listens to customer responses and cuts down on annoying verification questions as well as fraudulent behavior.
Pindrop has some mind-blowing capabilities. CEO Vijay Balasubramaniyan blew me away (and freaked me out a bit) by beginning a Q&A he and I did with his employees by playing a synthesized version of my voice greeting the group. His engineers pulled my voice from YouTube and fabricated some comments I never made. It wasn’t completely perfect, but I did recognize my voice.
Getting back to innovation happening outside Silicon Valley, Balasubramaniyan tells me his engineer pals in California rarely talk to him about engineering anymore. They’re preoccupied with issues of talent recruitment and retention and asking potential employees question like, “What coconut water do you need?” Pindrop, which has 330 employees (but doesn’t disclose it financials), has raised $122 million from the likes of Google Capital, Andreessen Horowitz and others. He says Georgia Tech provides ample hunting ground for engineers who aren’t as inclined to job hopping as their Silicon Valley counterparts.
Pindrop has the flourishes of a California company. Its office space is in a refurbished old building, just like in San Francisco’s South of Market neighborhood. Its conference rooms have clever names too: They are named for fraudsters Pindrop’s technology has caught.
Voice-recognition has been around for a long time, of course, led by listed companies like Nuance (nuan). Balasubramaniyan says old technology is based on rules fed into software, while his company’s artificial intelligence deploys deep learning. It can account for an aging voice, for example. Customers include banks, insurers, and retailers. He’s targeting automakers next, so drivers can speak commands to their car. (Chinese startup iFLYTEK talked about similar ambitions at the Fortune Global Forum last week in Guangzhou.)
It’s an exciting time to be in technology—in lots of different places.