Suki raises $90M to take its AI scribe beyond note-taking

Suki has raised $90 million in a Series E round, the company said Oct. 13. Its ambient AI documentation tool records conversations between patients and clinicians and converts them into notes for the electronic health record. Healthier Capital, Hercules Capital and Crosslink Capital came in as new investors. Every existing institutional investor returned as well: Venrock, Flare Capital Partners, Hedosophia, Philips Ventures and Zoom Ventures.

According to a news release, Suki's latest funding round will support the company's effort to build an open ecosystem for healthcare AI and will also fund its AI technology and integrations.

That second goal is the part a CMIO or health system CIO should read closely. Many of them bought an ambient scribe for one reason: to cut the time clinicians spend on documentation. Suki is now raising money on the idea that the scribe becomes the clinician's main way of asking the EHR, and outside reference content, for information. The central question is whether that holds up. If it does, the way these tools get bought and measured changes.

The timing makes the bet concrete. The round was announced five days after Suki launched Ask on Oct. 8, a feature that lets clinicians query patient records and Wolters Kluwer's UpToDate from inside the same app they use for documentation.

The growth behind the round

According to Fierce Healthcare's Cailey Gleeson, CEO and founder Punit Soni said the new capital will go to three areas: further work on the company's AI, growth of its commercial organization and a larger partner ecosystem. Soni told the outlet that, seen more broadly, healthcare AI is moving past documentation. He said health systems want more than a captured conversation. They want intelligence that helps with clinical decisions, lifts financial performance and makes operations more efficient.

Executives also gave numbers that explain why investors keep paying for this.

Executives also said revenue doubled in 2025 and is on track for similar growth this year. For a buyer, though, the partner figure may matter more than the customer count. More than 30 technology partners embed Suki's AI into their platforms. For operators, that suggests Suki's AI could show up inside tools they already use, not only in Suki's own app.

Healthier Capital partner Aman Mahajan will join Suki's board as part of the financing. In a statement, he said Suki's dual approach is what makes it particularly compelling. Suki sells its own app, Suki for Clinicians, and also powers AI features in platforms health systems already use. He also cited the company's deep clinical integrations and what he called demonstrated outcomes.

His larger argument is that AI's real opportunity in healthcare lies in rethinking "how work gets done across the health system," not in automating single tasks one at a time. His stated view is that the company could become basic infrastructure for healthcare's AI era. That's an investor's claim, not something the market has shown yet. But it tells IT leaders where the company plans to spend.

Inside Ask, across four EHRs

Suki points to Ask as one example of that vision at work. The feature is available across Epic, Oracle Health, athenahealth and MEDITECH, inside the Suki for Clinicians app. A clinician might ask for a patient's most recent A1C, the medications she is on right now, or whether any allergies are documented in her record. Clinicians can speak or type their questions in natural language, and Ask pulls up medications, allergies, vitals, lab results and chronic conditions, so they no longer have to dig through the EHR by hand.

Four-vendor coverage would matter most to a system running more than one of those EHRs across its sites, where a single question-and-answer layer could sit on top of different record systems. "Ambient AI started by helping clinicians document the encounter, but its potential is much larger," said Abhi Pathak, Suki's chief product officer, in the launch release.

The second layer is reference knowledge. Clinicians with UpToDate access can ask complex clinical questions and get answers grounded only in UpToDate's proprietary clinical content, which more than 7,600 physician authors, editors and peer reviewers develop and maintain. Responses show the sources behind the clinical evidence, include medication and dosing considerations, and link to supporting content and drug monographs. Put together, the two layers give clinicians one conversational interface for both patient-specific context and clinical knowledge.

The companies call Ask the first product integration of the UpToDate Expert AI API. The API lets healthcare technology partners embed source-grounded clinical intelligence in EHRs, ambient documentation platforms, clinical agents and conversational assistants, with flexible integration options for UpToDate users, source-connected outputs and controls designed for healthcare. Yaw Fellin, who runs clinical decision support and provider solutions at Wolters Kluwer Health, framed the goal as letting clinicians reach UpToDate answers "without having to leave their workflow."

Because that API is built for ambient documentation platforms as a category, the UpToDate grounding itself looks unlikely to stay unique to Suki. That suggests the differentiator Suki is betting on is the patient context it supplies alongside the reference answer.

The early evidence explains why Suki believes the scribe is the right place for this. In early evaluations, clinician evaluators chose UpToDate Expert AI responses that incorporated Suki-supplied patient context 83% of the time over responses without it (p<0.001). That preference held across clinical management, clinical safety, diagnosis, screening and dosing queries, and patient context was the primary driver. A scribe hears the visit, so it already has that context. A standalone reference tool doesn't.

The limits matter here. The 83% figure measures which answer evaluators preferred, not patient outcomes or error rates. The launch release does cite the American Medical Association, which ties extraneous cognitive load to 87% of individual medical errors, along with task-switching research showing that repeatedly shifting attention can reduce productivity, and Suki presents Ask as a way to reduce that load. But the release offers that as the reason for the product. It isn't a result Ask has produced.

Who owns the conversation

The industry, Soni told Fierce Healthcare, "doesn't need another collection of disconnected AI tools." He said health systems already juggle a surplus of point solutions that fail to communicate with one another, and those tools make work more complex when they should simplify it. What healthcare needs, in his view, is a connected, intelligent operating system that brings AI together across clinical, financial and operational workflows.

"The goal isn't to build every capability ourselves," he said, adding that what Suki envisions reaches well beyond ambient documentation or any one AI application. The release announcing the Ask launch describes a similar plan: push ambient AI past documentation so it can also support clinical reasoning, operational tasks and assisted revenue cycle workflows.

Teladoc Health shows the other model. On Oct. 7 it released SoloScribe, an ambient documentation tool that helps clinicians create structured clinical notes, inside Solo, its enterprise platform for hospitals and health systems. It shipped alongside SoloVitals, which measures heart rate and respiratory rate through patients' camera-equipped phones or tablets. SoloScribe is available now; SoloVitals is going to select customers ahead of a general release slated for December.

Suki's answer is its partner channel, where its AI runs inside other companies' products. Either way, scribing ends up living inside something else. What differs is which company provides the layer the clinician talks to. For systems that already run several AI point tools, that choice now belongs in the architecture review rather than in a department-level pilot alone.

The operational side of Soni's operating system is crowded with money, too. EliseAI, which automates patient calls, referrals, scheduling, insurance verification, chart prep and follow-up for specialty physician groups, raised $350 million at a $4 billion valuation, it said Sept. 29. That round was nearly four times the size of Suki's. It suggests that whoever aims to unify clinical, financial and operational AI will be integrating with well-funded specialists rather than replacing them.

A yardstick still being built

A product that does more also needs a broader way to measure it. In September, Forbes contributor Demetri Giannikopoulos wrote about Science at Suki and the Suki Research Collaborative, noting that ambient scribes have gained significant adoption this decade. Members of the network are Rush University System for Health, the Regenstrief Institute, MedStar Health's National Center for Human Factors in Healthcare, and the Office of AI in Medical Education at the University of Miami Miller School of Medicine. Its stated aims are to generate research, develop evaluation methods and build more consistent standards for measuring healthcare AI.

Among the collaborative's first projects, Regenstrief and Suki are building what the announcement calls a "gold-standard framework" for gauging how well ambient clinical intelligence performs and what impact it has. The goal is a set of standardized metrics and methods that cover its clinical, operational and financial impact.

According to Joshua Vest, Regenstrief's chief research officer, the goal is a common vocabulary more than a certification stamp. The work builds on LOINC, the international standard Regenstrief created earlier, which is widely adopted for identifying medical and laboratory observations. Vest said, "There are distinctions between adoption, implementation and successful outcomes. There's actually a pretty broad gap." There are, he added, "a lot of jumps and steps" between those stages of IT use.

Soni made a similar argument to Forbes. In his view, a verbatim transcript of every word spoken during a visit might be clinically worthless. The real objective, he said, is grasping the insight behind the conversation.

Ask makes that gap wider. You can check a note's accuracy against the visit it records. Judging whether a dosing answer was right for a specific patient is a different and harder test. If the Regenstrief framework comes out with metrics that cover answers as well as notes, health systems would have a shared vocabulary for writing these tools into contracts. If it stops at documentation, Suki's broader product will be ahead of any standard measure of it.