The startup QuantHealth claims it can predict how any patient will respond to any therapy, even novel ones

A pharmaceutical company can spend years and hundreds of millions of dollars learning that a Phase 3 trial was aimed at the wrong patients, the wrong endpoint or the wrong competitive bar. The venture-backed startup QuantHealth is trying to force that reckoning earlier.

QuantHealth says its AI clinical simulation platform can test trial designs before enrollment begins, model patient-level response and help sponsors decide whether a late-stage program is worth running at all.

Co-founder and CEO Orr Inbar said predicting how, say, a lung cancer patient might respond to a marketed immunotherapy is “the easy version.”

“What we really focus on is predicting how patients will respond to novel therapies,” Inbar said. “We help these companies take drugs that are still in phase one or even earlier and use our technology to simulate how real patients would react and respond.”

Why the bottleneck is moving downstream

Inbar positions QuantHealth downstream of the current wave of AI drug discovery, where Alphabet’s DeepMind and Isomorphic Labs helped turn protein-structure prediction into one of pharma AI’s most visible fronts, and cloud platforms such as AWS are pushing AI deeper into discovery workflows.

“Everything that’s happening in drug discovery is almost a foreshadowing of what’s going to happen in our space,” Inbar said.

Faster AI discovery pushes more candidate drugs into the clinical-development funnel, increasing pressure on trial design, patient selection and go/no-go decisions.

“A lot of the innovation and acceleration in drug discovery actually further exacerbates the bottlenecks in clinical development,” he said.

You’re going to get more and more drugs coming through the pipeline that need to be managed. —Inbar

AI-designed and AI-discovered drugs are entering the clinic. Insilico Medicine’s rentosertib, formerly ISM001-055, has provided a published Phase 2a proof point in idiopathic pulmonary fibrosis, while Generate:Biomedicines’ AI-engineered anti-TSLP antibody GB-0895 has moved into global Phase 3 studies for severe asthma. Inbar still described the current wave as early, with a handful of examples and Phase 1 throughput still short of a twofold increase.

“The tsunami hasn’t quite hit,” he said. “The AI-discovered drugs are just now making their way into the clinic. There’s a handful of examples, not enough to say that throughput is now 2x on phase one trials. We’re not there yet, but I do think it’s going to happen in the next few years.”

Simulate before you spend

QuantHealth’s pitch reaches upstream of conventional trial optimization. Inbar said the goal is to pressure-test the protocol before a sponsor commits to enrollment.

“What we do is help sponsors simulate the trial ahead of time, so they can understand whether the trial makes sense in the first place, whether they should run it at all,” Inbar said. “And if so, how to run it optimally: in a way that increases probability of success by selecting the right population, the right clinical presentation.”

He boiled QuantHealth’s simulation goal down to three operational questions: “Who to treat, when to treat, and how to treat.” That who/when/how rule also extends from protocol design to portfolio questions. Inbar said the same simulation technology can inform “what programs to advance further into the clinic, what programs to deprioritize, what drugs to in-license, what drugs to combine, what new indications to go after.”

From asset strategy to patient journeys

Inbar applied that frame at two scales: portfolio decisions about which assets to advance, in-license or deprioritize, and protocol decisions about which patients to enroll and how to define the comparator arm. A sponsor may need to compare assets, choose an indication, pressure-test a protocol or identify the patient subgroups most likely to drive a trial result.

QuantHealth’s public case studies describe using the platform to prioritize among three potential inflammatory bowel disease assets and to run thousands of in silico trials to define an optimal protocol design.

QuantHealth’s public materials cite a database of 350 million lives. Inbar illustrated the same scale with a U.S. example, describing a system that can create patient-level digital representations and test a therapy against a defined patient journey or subgroup.

“We basically have the entire U.S. population at our fingertips,” Inbar said. “We can take any person in the U.S., virtualize them, create a digital representation at any snapshot in time, and then run a virtual drug simulation on that person.”

In the interview frame, the mechanism comes down to patient-level digital representations built from de-identified claims and EHR data, then used to run virtual drug simulations across many possible trial designs. That helps the 350-million-lives claim read as a method for modeling patient journeys, beyond a database-size number.

What supports the claim

QuantHealth’s website states the claim directly: “Predict How Any Patient Will Respond to Any Therapy.” Asked about that line, Inbar stood by it.

“That’s bold, but I’ve been doing it for years now,” he said. “It’s a reality.”

QuantHealth points to commercial traction and partner distribution, while independent peer-reviewed validation remains limited. Accenture invested in QuantHealth in 2024, describing a platform trained on 350 million patients that can test thousands of protocol variations. Sanofi Ventures made a strategic investment in 2025, bringing total funding to $30 million. QuantHealth said in March 2026 that sales increased 8x year over year, its simulation portfolio had expanded to more than 600 trial simulations across 30 diseases and it had reported engagements with 12 of the top 20 global pharmaceutical companies. The company also said in 2025 that its technology would be offered on AWS Marketplace.

The public evidence mainly supports business momentum and market access. It gives less visibility into model accuracy for every patient-drug response.

From protocol to portfolio to capital allocation

QuantHealth also sells simulation as a tool for investors and business development teams. Inbar said investors have asked whether the company can help them evaluate biotech companies before major clinical readouts.

“If you can predict that phase three or whatever trial isn’t going to work the way you expect, or vice versa, you can make a better decision on how to allocate capital and what investments to make on that company or that drug,” he said.

That portfolio use case follows the same logic as the protocol use case: test strategic options before a sponsor spends on a trial, partnership or asset purchase. That could mean advancing a program, deprioritizing one, in-licensing a drug, combining assets or looking for a new indication.

Accenture gives QuantHealth a consulting channel into pharma R&D transformation work. AWS gives it cloud distribution and a familiar environment for sponsors already building clinical data infrastructure. Sanofi Ventures puts the company close to a drugmaker managing the same pipeline and portfolio questions QuantHealth wants to influence.

The startup is betting that clinical development can be simulated early enough, at enough scale and with enough patient-level resolution to change which trials, partnerships and asset purchases move forward. “All of this opens up new worlds of possibility for how we even think about clinical trials,” Inbar said. “I think we’re potentially opening up totally new regulatory pathways.”