Raleigh AI startup Pryon raises $100 million
Add Axios as your preferred source to
see more of our stories on Google.
Illustration: Sarah Grillo/Axios
Pryon, a Raleigh-based artificial intelligence startup that makes a virtual assistant for companies, raised $100 million from investors in its most recent round of funding.
Driving the news: The investment, announced Tuesday, was led by the US Innovative Technology Fund — an AI-focused fund chaired by the billionaire investor Thomas Tull that has invested in companies like Anduril and Shield AI.
- The investment values the company at between than $500 million and $750 million, according to a person familiar with the matter.
Why it matters: Interest in AI startups has increased rapidly since the last year's release of ChatGPT, and Pryon's raise is the most significant funding bet on a Triangle-based AI company.
- The company is headquartered in Raleigh Founded's coworking office in Gateway Plaza. Its employees, which number over 100, are spread throughout the country.
Zoom in: Pryon was founded in 2017 by Igor Jablokov, whose previous startup Yap was acquired by Amazon and used to create the Alexa.
- Jablokov has created an AI assistant that lets businesses upload huge amounts of internal information and data — and then will answer questions from workers related to the information.
- In the past year, Pryon has begun building out a sales team and has landed customers looking to use it as a productivity tool in the finance, government and industrial spaces.
The big picture: Jablokov told Axios there will be a lot of AI companies that rise and fall in the space in the coming years.
- He noted that the team behind Pryon has been working in the field for years and is not jumping on a fad.
What they're saying: Jablokov said the key for Pryon is to create an AI system that can be trusted by companies to give answers based on real sources.
- "Every answer coming out of Pryon is always anchored to ground truth," he said. "It always shows you the exact page and highlights exactly where it learns something from."
- "In order for people to trust output from an AI, they need to trust fundamental sources where this comes from," he added. "Nobody wants anything hallucinating on the flight deck of a 747."