Ambience Healthcare, an artificial intelligence company for healthcare organizations, said Tuesday it received $70 million in a Series B funding round.
The round was co-led by OpenAI’s Startup Fund and venture capital firm Kleiner Perkins, with participation from existing investors Andreessen Horowitz and Optum Ventures. OpenAI is the company behind the publicly available generative AI large language model ChatGPT, which has had the healthcare industry buzzing since its arrival in November 2022.
Read more: Happy birthday, ChatGPT: AI tool still generating buzz, concern
Ambience, founded in 2020, has developed an AI operating system that uses the technology for clinical note-taking and coding as well as generating referral letters and pre-visit and after-visit summaries. Co-founders Michael Ng and Nikhil Buduma met at the Massachusetts Institute of Technology.
Through their connections, Ng and Buduma were able to observe the early days of OpenAI and work with Jeff Dean, Google’s chief scientist and a leader of the big tech company’s AI efforts, Buduma said.
“We just had this front row seat to this hyper evolution of the space,” Buduma said. “By 2020, I think we saw the opportunity if the right team could come together to fine-tune these models specifically for medicine with the right safety guardrails embedded into the workflow. The most capable models could create a ton of impact in healthcare.”
Ng said the $70 million will be used to accelerate Ambience’s product road map and expand its team. With this round, Ambience has raised $100 million.
Ambience's customers include Walnut Creek, California-based health system John Muir Health, Houston-based Memorial Hermann and the University of California at San Francisco Health. John Muir Chief Medical Information Officer Dr. Priti Patel said the health system was looking for an ambient AI solution that could listen to clinician-doctor conversations and then generate notes automatically.
“The president of our medical group and I had tested the [AI tool] with a 20-minute complicated patient scenario,” Patel said. “When we finished, they pressed a button and it generated this incredible note.”
The technology has been easy to implement but the patient consent process has required some training, Patel said. The system won’t use the tool if it doesn’t get consent from patients and Patel said some patients are uncomfortable with AI.
