Cardinal drafts a complete, ICH M11-ready protocol from your Investigator's Brochure — then checks your design against the precedent of past trials, so you catch the eligibility, endpoint, and enrollment risks before they become amendments.
Nearly two-thirds of trials miss their primary objective — and the cause is often a design choice made weeks earlier: eligibility criteria too narrow to enroll, endpoints regulators question, schedules sites can't realistically run. By the time these surface, the fix is a costly amendment. The risk is visible early — if you have something to compare against.
Restrictive eligibility, mismatched endpoints, and infeasible schedules rarely announce themselves on the page. They surface during enrollment or regulatory review — when correcting them means an amendment.
Background, rationale, safety monitoring, and statistical frameworks are functionally identical across protocols, yet manually rewritten for each new study — weeks of effort that add no design value.
What worked — and what failed — lives across hundreds of thousands of past trials. Until now, that evidence has been inaccessible at the one moment it matters most: while you're still designing the study.
Upload your Investigator's Brochure and provide your study inputs. Cardinal returns a complete, ICH M11-aligned protocol draft — schedule of assessments, statistical framework, and safety monitoring intact — and benchmarks your design against the precedent of past trials in your indication.
Cardinal is built around the workflow your medical writers already know — not a new paradigm to learn.
Drop in your Investigator's Brochure. Cardinal extracts mechanism, dosing, PK/PD, and safety data — verbatim, never fabricated.
Twenty-five structured inputs covering phase, design, blinding, randomization, endpoints, sample size, and population.
Editable Microsoft Word output, formatted to ICH M11, with section structure, SoA tables, and statistical framework intact.
Cardinal does not generate protocols from a blank slate. Every endpoint suggestion, design choice, and safety framework is grounded in the precedent of similar trials — surfaced from a structured index of past clinical research.
When ACR20 was the standard primary endpoint for Phase 2 RA trials, Cardinal would suggest it. Today, ACR50 has overtaken it — and Cardinal reflects that shift, with the data to back it up.
This is the difference between an AI tool that writes for you and an AI platform that thinks alongside you.
Cardinal is not a writing assistant. It drafts the protocol, then helps you pressure-test the design — built from first principles around how clinical research teams actually work.
Mechanism, dosing, PK/PD, and safety pulled verbatim from your uploaded Investigator's Brochure. Never fabricated, never approximated.
Available nowCardinal surfaces endpoint patterns, statistical approaches, and design benchmarks from past Phase 1–3 trials registered on ClinicalTrials.gov.
Roadmap · Q3 2026Catches cross-section contradictions before they reach review — treatment duration vs. SoA, endpoint definitions vs. statistical sections.
Roadmap · Q2 2026Output is a properly formatted, section-structured .docx — not raw text. Drops into your existing review workflow from day one.
Available nowMulti-user review, version control, and threaded comments for your full authoring team. No more emailing Word files between fourteen people.
Roadmap · Q3 2026Protocol structure, section ordering, and terminology aligned to the ICH M11 guideline — the current international standard for clinical protocols.
Available nowMerquell is an AI company building authoring tools for the world's most consequential documents — clinical trial protocols, regulatory submissions, investigator's brochures.
Today, these documents are still authored in Microsoft Word, passed around by email, and reviewed through processes built on tribal knowledge accumulated over decades. We exist to change that — not by replacing medical writers, but by amplifying them with AI built to the rigor of the science they produce.
Cardinal is our first product. Built from first principles, with input from medical writers and clinical operations leaders. We start with protocol authoring — and build toward a future where every step from IND preparation to regulatory submission has proper AI-native tooling.
Merquell Technologies Inc., federally incorporated in Canada. Headquartered in Toronto, Ontario.
Pre-launch. Working with select design partners to validate Cardinal and shape the roadmap. Not yet generally available.
Series A–C biotechs (20–150 employees) and clinical research organizations requiring Phase 1–3 protocol authoring.
ICH M11 compliant generation. US FDA, EMA, and Health Canada awareness built into the core architecture.
Clinical research runs on trust and traceability. Cardinal is engineered for the security, accuracy, and auditability that regulated work demands — so the output stands up to the people who review it.
Your Investigator's Brochures and protocol inputs are never used to train any AI model, and are accessible only to your organization. Confidentiality is the default, not a setting.
Your IB and protocol inputs are encrypted in transit and at rest, processed under strict data agreements with the enterprise AI providers that power generation.
Every generated section traces back to its source — IB extract, precedent reference, or user input. Designed for regulatory traceability from day one.
Built with awareness of FDA, EMA, PMDA, and Health Canada regulatory frameworks. Generation logic adapts to regional submission requirements.
We are working with a select group of design partners — biotechs and CROs who want to shape an AI-native protocol authoring platform from the ground up. Request a walkthrough and we will be in touch within twenty-four hours.
No sales pitch. A thirty-minute walkthrough with the founder.
We will be in touch within twenty-four hours to schedule your walkthrough. Watch your inbox for confirmation.