How Synthflow AI is cutting through the noise in a loud AI voice category

The conversational AI market has exploded since ChatGPT was released in November 2022 and is predicted to grow into a nearly $50 billion global industry by 2031, according to MarketsAndMarkets.

Synthflow AI is just one of many companies building in this space that hopes to stand out from the pack because of its focus on being enterprise-grade and easy to set up.

Berlin-based Synthflow is a no-code platform that lets enterprises build and deploy customized white-labeled voice AI customer service agents. The company, which launched in 2023, has amassed more than 1,000 customers and has handled more than 45 million calls.

The startup’s voice agents are both HIPAA and GDPR compliant and can be plugged into more than 200 integrations with other enterprise platforms including Salesforce, Twilio, and HubSpot, among others.

Hakob Astabatsyan, co-founder and CEO, told TechCrunch that he and his co-founders, Albert Astabatsyan, now CPO, and Sassun Mirzakhan-Saky, now CTO, started messing around with OpenAI’s ChatGPT API back in early 2023 to find potential ways to build no-code business applications on top of the AI model.

They started with a text-to-text AI bot and then tried to build a voice bot. When they realized how much harder voice was, they got excited about the potential.

“We realized, oh my god, voice is really complicated, right? To actually make AI speak in real time like we do, having this 400 milliseconds latency, and handling interruptions, it turned out to be like such a complicated task,” Astabatsyan said. “We fell in love with this problem, and we said, look, we’re gonna work only on voice bots from now on.”

The group formed Synthflow and spent the rest of 2023 building and launched its first version of the product at the beginning of 2024 before releasing an enterprise-grade version of the tech at the end of the year. The company grew 15x last year and has seen over 90% retention from its enterprise customers, according to Astabatsyan.

“We process 5 million calls monthly,” he said. “Last year, it was like, I don’t know, 1 million, 2 million, and and then we started growing very quickly. This is where Synthflow started really getting better and better because we had this velocity.”

The startup also recently raised a $20 million Series A round led by Accel with participation from existing investors Atlantic Labs and Singular. Astabatsyan said the company raised this recent round so that it could expand its team, boost research and development, and open its first U.S. office in an undecided location.

Luca Bocchio, a partner at Accel, told TechCrunch that the Accel team had been tracking Synthflow since it started developing its first product. What stood out to Bocchio was the founding team’s drive and its early push into building enterprise-friendly integrations.

“This team has [had] really strong views since the getgo about creating more depth with the technology and extensive integrations across CRMs, across tools enterprises may use to really provide enterprise-grade compliance,” Bocchio said.

Regardless of the company’s traction, conversational AI seems poised to be a tough category. There are numerous other companies building in the space including Bret Taylor’s Sierra, which has raised $285 million in VC money, and Bland AI, which has raised more than $50 million in venture funding, to name a couple.

“AI is moving so fast, and sometimes things happen faster than you would expect,” Astabatsyan said. “But for us, it’s very clear. We’re at this stage where, I would say, [we’re] in a post-product-market-fit era, where we know who our customers are. We have a pretty clear idea what’s our product roadmap, and where we want to be in the next three to five years.”