Intake AI Raises $2.5M for Clinical Trial Operations

Keeping a paper version is a telling product choice for Intake AI. The clinical-research startup builds source documents that can travel into electronic systems or be printed for the people running a study. It has raised a $2.5M pre-seed to develop AI agents for trial study start-up and operations, according to co-founder and COO Miles Wingfield's October 13 announcement.

Kapor Capital led the round, with Blankspace Ventures and Act One Ventures participating. CEO Samir Hooker, CTO Lakshman Peri and Wingfield are building around the administrative work that sits between a study protocol and a research team's ability to use it. Their choice to work with existing documents and software gives the financing a concrete destination: the coordinator's desk.

The company reports live deployments at research sites across the country, without disclosing a site count or revenue. Its first live workflow produces source documents, while additional administrative workflows are in development. For clinical-operations buyers, that creates a useful way to examine the product: follow a piece of study information through preparation, review and entry into the system where the next person needs it.

A protocol still needs somewhere to land

A study protocol describes what a trial requires. Intake AI's source-document workflow turns that material, including the schedule of assessments, into documents for individual visits mapped to electronic case-report-form requirements. The company supports electronic delivery and printable output, an accommodation that keeps the site's working format in view.

That last detail deserves attention. A coordinator can have a perfectly good document and still face another task getting it into the right application. Intake AI's announcement describes placing approved forms into existing eSource systems, making the downstream destination part of the build. The product's value proposition therefore reaches beyond generating a first draft: it includes reducing the transfers around that draft.

Research teams retain control, according to Wingfield's announcement. That allocation of responsibility fits the work being offered, because the team using a visit form must understand what it contains. A generated field, its mapping and its destination are all part of the same operating task, even when different software screens separate them.

The workflow also gives the startup a specific place to begin expanding. Source documents are available today, while the company describes further workflows as still in development. The $2.5M funds an operations ambition with a defined first deliverable; the public announcement provides no detailed allocation of the capital or timetable for additional releases.

Three founders came through clinical research

The official leadership page identifies Hooker, Peri and Wingfield as Intake AI's three co-founders. Hooker is CEO, Peri is CTO and Wingfield is COO. Kapor Capital's portfolio independently names all three, lists the company's founding year as 2025 and records its investment year as 2026.

The founders' route into the subject gives the product choice some weight. In a 2025 UC Berkeley interview, Hooker connected his earlier company, Vitify Health, to his mother's chronic illnesses and the difficulty of reaching useful trial opportunities. Intake AI's team page now describes his earlier work matching patients with trials and his current focus on the operations performed at research sites.

Wingfield's funding announcement describes another family encounter with the same industry: his father suffered a serious stroke shortly before Wingfield moved to college. A trial was considered, and Wingfield saw research staff spending time locating documents and requesting system access. Those are his account and his motivation; the announcement supports the connection without needing a fictional scene around it.

The team page describes Peri as a UC Berkeley graduate and published researcher with experience at Berkeley AI Research and the Tata Institute of Fundamental Research. Together, the founders have chosen a particular unit of work to automate. Their backgrounds explain why trial-site administration has their attention, while the software still has to earn its place in each site's daily routine.

Answers and actions share the study record

Intake AI's product page describes a suite combining document questions, system connections and workflows. Chat draws on material such as protocols, standard operating procedures, budgets and staffing records. The stated goal is to answer questions whose evidence lives across several files, with sources that a user can examine.

That creates a different relationship with the coordinator than a standalone writing box. A document assistant may help someone understand the study; the source-document workflow is intended to carry that understanding into work the study needs. Intake AI presents the suite as sharing the same document corpus and connections, a design choice that places retrieval and execution beside one another.

Its Chrome extension privacy policy, dated September 10, 2026, describes the mechanics of authorized browser actions. The extension can inspect supported pages, populate study information and read values back to check completion. It also describes pausing when uncertainty or mapping ambiguity calls for human review, and allowing users to stop or correct the work.

Those are company-described capabilities and controls. Public materials reviewed for this announcement provide no independent comparative evaluation of deployment outcomes, so speed claims and clinical benefits remain outside this report. The useful operating question is specific enough without a performance slogan: can a research team inspect the evidence behind an answer and then review the work that answer helps produce?

The purchase lives inside the next study

Clinical research creates several kinds of administrative demand. DevCuration's coverage of clinical trial finance concerns another part of that operating environment, while its report on medical certification follows the institutions around bringing medical technology to market. Intake AI's focus is the site's study-preparation and information work.

For a buyer, the relevant scope is the supported workflow and its destination. Source-document preparation, questions across study materials and approved eSource entry offer concrete tasks to inspect during a demonstration. The printable route also makes adoption legible to teams whose study records still include paper, without assuming that every trial uses an identical software stack.

The pre-seed syndicate has financed the startup's stated effort to develop those agents. The announcement leaves valuation, prior financing history, customer counts and financial results undisclosed. That keeps the commercial picture narrower than the founders' broader ambition, which includes giving research teams more time for the work that brought them into clinical research.

A new study arrives with fresh documents and its own visit requirements. Intake AI is building toward that recurring arrival, where the coordinator must translate the protocol into an afternoon's usable work. The next workflow the company ships will join that same desk, alongside the forms and systems the site already trusts.