Time's almost up! There's only one week left to request an invite to The AI Impact Tour on June 5th. Don't miss out on this incredible opportunity to explore various methods for auditing AI models. Find out how you can attend here.
Nabla, a French startup building an AI copilot to accelerate how doctors work with patients, today announced it has raised $24 million in a series B round of funding.
The investment comes from Cathay Innovation and Zebox and will be used to build out Nabla’s offering and accelerate its rollout in the U.S. market. The funding takes the total capital raised by the company to $44.6 million.
There have been a lot of copilots, although their adoption in the healthcare sector has been slow due to the concerns of doctors and clinicians. With its AI copilot, Nabla appears to have broken that barrier, giving doctors a helper that sits on their desk and produces the desired results.
What exactly does Nabla help with?
Nabla Copilot sits as a web app or Chrome extension on a doctor’s computer and listens to their consultation with patients. Then, using the encounter and the magic of AI, it generates clinical notes almost instantly, saving doctors the trouble of documenting the information manually.
June 5th: The AI Audit in NYC
Join us next week in NYC to engage with top executive leaders, delving into strategies for auditing AI models to ensure fairness, optimal performance, and ethical compliance across diverse organizations. Secure your attendance for this exclusive invite-only event.
These notes carry all the necessary information from the consultation, from the patient’s issue to their medical history and the medicines/tests prescribed to them. This information can then be reviewed/edited by the physician and filed into an electronic health record. The platform is also on track to get a feature that uses the discussion and examination to generate a set of patient instructions, on behalf of the physician. This letter can be saved as a PDF and printed for the patient.
To bring these capabilities to life, Nabla uses real-time speech-to-text AI. The model, a combination of Microsoft’s off-the-shelf speech-to-text API and fine-tuned Whisper model, generates the transcript of the conversation. This transcript is converted into a summarized clinical note using the GPT series of large language models (LLMs).
Nabla says all the personally identifiable information (PII) is masked when the LLM processes the transcription and unmasked once the output has been produced. Plus, neither the speech nor the summarized notes are stored, unless explicitly consented to by the physician and the patient.
Significant impact in less than 10 months
Historically, physicians have had to spend up to 40% of their day documenting consultations. With Nabla, the problem is just gone. In just 10 months of its launch, the copilot has been adopted by more than 20,000 providers. The company also claims that it will soon hit three million consultations.
“With Nabla, I am finally getting my weekends back. What took 8-10 hours now takes maybe 3-4 hours. Better yet, because I no longer need to be a scribe, I get to actually be a psychologist and focus on my patients! After only a couple of months, I can’t imagine my practice without Nabla,” M. Tursich, a psychologist at Mankato Clinic, said in a statement.
The company claims that the fine-tuned models provide highly accurate outputs, with only 5% of the notes requiring adjustments. Our tests on the web app also provided similar outcomes. This is expected to get even better as the models improve in the near future.
Nabla plans to use this funding to expand the reach of its copilot in the U.S., but its ultimate goal is to build an AI-driven future for physicians. The company aims to build the “most intuitive and reliable ambient AI” to support them across the clinical spectrum up until medical decision support. It will also use the fresh capital to launch additional language options for the assistant.
Currently, Nabla Copilot can be accessed by any email, not just the official provider email, and supports three languages: English, Spanish, and French.
