Before Parloa was a USD 3bn company, it was a consulting shop that built voice apps for Alexa. In 2016 Malte Kosub and Stefan Ostwald started Future of Voice, which Kosub says was one of the first companies in Germany to build Amazon Alexa applications. Nobody was sure yet that talking to machines would be more than a party trick.
Kosub came from Hamburg, and studied business administration at the University of Hamburg until 2016. Ostwald shared his obsession with voice technology. They had no giant research lab behind them, only timing: speech-to-text was improving fast, and brands wanted to be on the new devices.
Their choice was to be voice-first and AI-ready from the beginning, while most competitors had begun with chat.
Consulting paid the bills. It also showed them something. Working with large companies, the pair noticed a gap in serious technology for customer contact centres, the place where a company's patience is tested a thousand times a day. One profile says they were building software on the side while consulting kept the lights on. In 2018 they founded Parloa.
Their choice was to be voice-first and AI-ready from the beginning, while most competitors had begun with chat. Kosub has said he was convinced by early speech breakthroughs that talking is a more natural interface than typing. It was a bet that looked early, then looked lucky.
The hardest thing, he has said, was not the algorithms but people. Finding hires who combine skills and cultural fit can be challenging, a quiet admission from a founder whose product is supposed to replace queues of agents.
The money followed the product. A USD 66M Series B led by Altimeter arrived in April 2024. In May 2025 a USD 120M Series C valued the company at USD 1bn. Then, in January 2026, General Catalyst led a USD 350M round at USD 3bn, tripling the valuation in about eight months. The company now has offices in Berlin, Munich, New York and London, with a reported headcount in the hundreds.
The lesson is easy to miss. Parloa's edge was built while it was a service business, learning customer problems at close range and writing software only once the gap was clear. Consulting is not a detour, it can be the research phase.