What a Real AI Breakthrough in Healthcare Looks Like

Motive Notes，Motive Force 支持那些看到隐形真理的创始人，并与他们合作直到这些真理无法被忽视。关注

Today HealthLeap announces a $38M Seed and Series A led by Sequoia Capital, First Round Capital, and Hummingbird Ventures. I’ve been a partner since seed, working alongside founders Jemima and Josiah. This is the story of what I saw in them, why their approach to healthcare AI is different.

The Risk Hospitals Can’t See

The breakthrough is making invisible clinical risk economically capturable without changing clinician behavior.

Take malnutrition, the most under-captured condition in the hospital. Roughly a third of inpatients are at risk of it. Fewer than 9% get diagnosed. At a typical 1,000 bed hospital, that means roughly 5,500 malnourished patients missed every year. The intervention is low risk, the reimbursement is meaningful, and outcomes improve when it’s caught early. No one is negligent; a clinician under time pressure can’t continuously scan labs, notes, vitals, medications, and demographics across a full inpatient census. Software can

HealthLeap layers intelligence on top of what already exists. They ingest messy EHR data: labs, notes, vitals, medications, demographics, and continuously surface high-risk patients, writing a nutrition-risk score into the dietitian’s EHR flowsheet each morning. In the peer-reviewed validation at Cedars-Sinai, across 166,841 admissions over nearly four years, the model was 49% more sensitive on the first day of hospitalization than the nurse-administered screen in use, and flagged malnutrition a mean of four days before the first dietitian documentation. HealthLeap measured the downstream result at Cedars-Sinai: 39% more diagnoses with the same staffing, worth about $11M a year. No added staff, no new alert piled onto an overloaded system.

Hospitals are overloaded systems. Clinicians make judgment calls based on a handful of data points under time pressure. They focus on the urgent case in front of them. The subtle signals: gradual weight loss, appetite decline, lab drift, notes buried in documentation, disappear into noise. What’s missing is continuous visibility. HealthLeap provides it.

How I Found Them

In July 2022, I received an email from JP Posma, founder of Zaplib:

“Josiah, meet Oana. She’s a VC and on the side also likes to help female engineers and founders shesready2.ai… I immediately thought of the inspirational story of how your sister came up with the idea for HealthLeap.Oana, HealthLeap is a super inspiring startup and Josiah is an amazing human.”

JP introduced me to Josiah so he could introduce me to his sister and cofounder Jemima, and she could join She’s Ready to AI, the community I run for women in tech, because the most interesting founders often exist before the market notices them. When Jemima and I first spoke, she was calling from inside a large hospital system in Los Angeles, where she was working on-site. That detail stayed with me. Hospitals do not casually grant access to their EHR systems.

Jemima cared deeply about patients. What struck me most was her unusual desire for scale, and it wasn’t accidental: she grew up with the idea that if you are going to do something, do it in a way that changes many lives, not a few. That drive is what I call motive force, the inner drive that sustains motion through uncertainty, and it’s why I named my firm after it. Over the life of a company, the market, the product, and the plan all change. The founder’s motive force doesn’t.

Jemima later brought her brother Josiah to one of the builder events I organized, and I remember watching them interact. Both of them carry a quiet certainty, with no arrogance and no hype behind it. Josiah brings mathematical intensity and competitive sharpness. In JP’s words, “Josiah is an amazing human.” Together, the work is rigorous and grounded.

The Pattern Institutions Don’t Reject

The lesson from HealthLeap has nothing to do with malnutrition. Don’t fight the immune system of the institution: find where intent and economics already align, and build where amplification is welcome before autonomy is tolerated.

That’s the pattern I back at Motive Force. I invest early in founders like Jemima and Josiah who build technically rigorous products designed for adoption from day one, AI that fits reality so precisely that adopting it feels obvious.

Software that helps clinicians see will do more for healthcare than software that tries to replace them. When a product improves outcomes and pays for itself, with no behavior change required, hospitals stop treating it as an experiment.

HealthLeap’s product fits. In a hospital overloaded with alerts and dashboards, that restraint is rare. Beautiful software in healthcare is not about visual polish. It’s about calm leverage.

Join HealthLeap

If you’re an engineer or clinician who wants to build products that save lives while respecting how healthcare works, HealthLeap is hiring. They want people who understand that the best AI makes institutions better at what they already want to do.

This team sits in hospitals and earns trust before asking for access. Clinicians use the software because it makes their jobs easier.

You can learn more at https://careers.healthleap.ai/.

Build with Motive Force

If you’re a founder, or someone who hasn’t started yet but feels that inner drive, I want to hear from you.

I met Jemima and Josiah long before I wrote a check, and we worked together before there was a formal partnership. Trust preceded capital. That’s how I like to invest, and it means it’s never too early to reach out, even at the imagination phase. Partnership starts with shared thinking, long before a term sheet.

I’m looking for founders who build rigorous products designed for adoption from day one, and who have the motive force to keep moving through uncertainty.

You can reach me at oana@motiveforce.ai.