Taxo.ai Inc., a company building an artificial intelligence reasoning engine for healthcare administration, said today it raised $5 million in seed financing to reduce the burden on physicians so that they can spend more time on patient care.

The funding round was led by Y Combinator, General Catalyst and Character, with participation from prominent angels including Zeus Living founder Kulveer Taggar and Rocket Money founder Yahya Mokhtarzada.

Founded by Dr. Ahmed Kerwan, chief executive of Taxo, a practicing physician himself, Taxo addresses a critical pain point in healthcare by using AI to automate administrative tasks that currently consume a significant portion of medical professional’s time. The company’s approach aligns with the 2025 trend of developing intelligent agents capable of handling complex work.

The name of the company comes from the word for “taxonomy,” the branch of science that deals with classification. It’s also related to classifying medical claims if they were acceptable, rejected or needing more information; one of the prototype iterations of the software that would evolve into Taxo. However, “taxonomy.com” was too expensive, so Kerwan opted to buy “taxo.ai” for only $100.

“First and foremost, we started by creating an AI that integrates with electronic health records to automate all the paperwork specifically in payer and provider communications — so, prior authorization, letter, claim, adjudication, coding, billing,” Kerwan told SiliconANGLE in an interview.

Most doctors are often reluctant to use tools they consider to be a “black box,” unless these tools provide clear reasoning for their decisions and connect all outputs to the original records. To address this concern, Taxo ensures that all its results are auditable, which is intended to build trust with healthcare providers.

“It mirrors all the data from your existing software in one place, in a standalone app,” explained Kerwan. “Then you can have it do a number of communication and paperwork-related functions.”

The AI software also ensures data security and regulatory compliance including the Healthcare Insurance Portability and Accountability Act.

Kerwan explained that right now there’s such a paperwork burden that sometimes he has had days where he had three hours of seeing patients and then six to seven hours of doing paperwork. It’s not always doctors who are burdened by paperwork. For example, there are back-office staff hired or outsourced specifically to do prior authorizations — a procedure where health insurance plans require a doctor’s office to get approval for particular services or medications.

“We just want to help reduce the need for all this excessive hiring and all these excessive hours being spent on paperwork,” Kerwan said.

Under the hood, Taxo doesn’t use a specialized foundation AI model, instead, it uses a model-agnostic approach through a unique “meta-layer” that sits atop a foundation model. That means Taxo can incorporate multiple foundation models and isn’t locked into any single AI technology and can fine-tune for specialized healthcare administration tasks.

Kerwan explained that this also avoids having to train and fine-tune a foundational model, which would be extremely expensive and take a long time to bring to market each time. It would also allow the company to adapt rapidly alongside emerging AI technology.

“We can’t compete against Anthropic or OpenAI, but we also want to get better as AI gets better,” Kerwan said.

Now that Taxo has secured funding, Kerwan said that the company will build out the AI meta-layer technology and to do that it intends to focus on hiring AI talent. Another focus of the company will be partnering with healthcare institutions by forming a team on the ground that will act as anthropologists to help build custom solutions.

“My first degree was actually in anthropology before medical school,” Kerwan said. “I think it’s not just the technology that matters; it’s also understanding the culture of the enterprise, how they think, how they work, the day-to-day.”

Image: SiliconANGLE/Microsoft Designer