Aevius launches with $22 million to make medical data computable
Aevius launched with $22 million led by Insight Partners to build AI models that turn fragmented medical records into refreshable, multimodal patient views for hospitals, insurers and trial sponsors.
By Nathan Brooks
4 min read

What's News
- Aevius launched with $22 million in funding led by Insight Partners.
- The Mass General Brigham spinout was founded by Rowland Pettit, Ming Yang "Max" Lu, Faisal Mahmood and Long Phi Le.
- Founding team members previously helped build Modella AI, which AstraZeneca acquired.
- The capital funds model and product development, commercial deployment, and hiring across engineering, clinical and commercial teams.
- Eric Topol chairs Aevius's Scientific Advisory Board; Chris Coburn is chief innovation officer at Mass General Brigham.
Aevius has launched with $22 million in funding led by Insight Partners. The Mass General Brigham spinout will use the capital to build AI models that convert fragmented medical information into clinical intelligence for healthcare providers, insurers and life sciences companies.
Founded by Rowland Pettit, Ming Yang "Max" Lu, Faisal Mahmood and Long Phi Le, Aevius plans to spend the money on model and product development, commercial deployment, and expansion of its engineering, clinical and commercial teams. Members of the founding team previously helped build Modella AI, which AstraZeneca acquired, and they are now applying that experience to structured, unstructured and specialty multimodal medical data.
The fragmented data problem
Healthcare organizations have spent decades digitizing clinical information, and the resulting data stays distributed across different systems, formats and medical vocabularies. Some details sit in structured fields; others live in physician notes, pathology reports, medical images, laboratory results and specialty systems.
Linking those sources for research or clinical applications can require substantial manual preparation before any model runs. Aevius develops models that learn relationships across data types and across a patient's longitudinal history, producing what the company describes as a refreshable multimodal patient representation.
The representation is intended to give a computable picture of a person's clinical condition and trajectory, and to update as new information arrives. The practical gain is reuse: organizations can build multiple applications on the same representation rather than rebuilding a data pipeline for every new question.
What can buyers build on it?
Aevius splits its targets into three groups, with different jobs for the same underlying model:
- Health systems: clinical research, care delivery, resource planning, and earlier identification of patient needs.
- Health insurers: longitudinal clinical intelligence to track changing patient requirements and support more coordinated intervention.
- Life sciences: studying disease progression, evaluating treatment outcomes, identifying potential biomarkers, and finding patients who may meet clinical study criteria.
The company also expects its patient representations to power AI agents. It argues many existing agents assemble an understanding of a patient by retrieving isolated pieces of text for each question, whereas Aevius aims to give those tools a persistent representation grounded in the broader medical history.
That shift matters for anyone building clinical tools now. Researchers, care teams and agents can take on new questions without reconstructing the patient's record each time, working from the whole patient rather than fragments.
What the founders and backers say
"Our mission is to make medicine computable. As a physician-scientist, I've worked across many types of clinical data, and there is far more knowledge in those records than we're able to use today. We're bringing general medical intelligence to that data, so organizations can build on a deeper understanding for their patients." — Rowland Pettit, co-founder and CEO of Aevius
"Most AI agents today piece together their picture of a patient from text, one question at a time. We're building a native representation of each patient's multimodal, longitudinal history that powers prediction and analysis across an entire population. Researchers, care teams and AI agents can take on new questions without starting over, working from the whole patient rather than fragments of a record." — Ming Yang "Max" Lu, co-founder and chief technology officer of Aevius
"The value of medical AI will be measured by how well it promotes accuracy. Aevius is pursuing that goal with models grounded in the breadth and complexity of patients' medical histories." — Eric Topol, chair of Aevius's Scientific Advisory Board
"Mass General Brigham develops AI with the ambition to improve healthcare far beyond our own hospitals. Aevius was spun out of MGB to bring together research developed here and a team with experience taking medical AI to market. We expect that organizations throughout care delivery and life sciences will soon benefit from this powerful technology." — Chris Coburn, chief innovation officer at Mass General Brigham
"As with other industries, AI in healthcare is only as valuable as the data that drives it, and clinical data is among the hardest to unlock. Aevius is a rare combination of talent and technology. We are thrilled to back these inspiring founders and their vision to make medicine computable, deliver clinical intelligence to the enterprise, and ultimately improve outcomes for patients." — Dylan Morris, managing director at Insight Partners
What does this change for hospitals, insurers and trial sponsors?
Practically, it collapses a manual data-assembly step into one shared layer. A hospital that already holds decades of digitized records gets a single refreshable patient representation to feed research, care planning and resource allocation. Insurers get a longitudinal view for coordinated intervention. Trial sponsors get a way to locate eligible patients and track disease progression.
The unanswered piece is adoption outside the founding hospital system. Aevius has not disclosed pricing, customer contracts or deployment timelines, and Insight Partners' capital now has to move the company from research and early development into broader commercial deployment.
The next test is commercial rather than scientific: whether care organizations buy the representation instead of assembling their own pipelines will decide if the $22 million turns into a durable business.
Source: pulse2.com
More from Nathan Brooks
Show full bio
News editor covering regulation, policy and their impact on entrepreneurship, startups and venture funding at Business Bearings.
2,426 articles