Built for clinical trial sponsors

Turn a clinical protocol into a validated, IRB-ready recruitment strategy — in hours, not weeks.

Horizon AI is the feasibility and study-startup platform for pharma and biotech sponsors. Validate enrollment assumptions, model recruitment risk, and generate IRB-ready materials — before a single site is activated.

Upstream of CROs Sits before site activation and vendor engagement
Audit-logged Every output versioned, citable, and IRB-ready
Vendor-neutral Hands off to your existing execution partners
Core capabilities

From protocol PDF to IRB-ready execution package

Four AI layers that turn a clinical protocol into a defensible, sponsor-grade recruitment plan — with the math, the materials, and the audit trail.

1

Protocol ingestion & structuring

Upload a protocol synopsis or full PDF. AI extracts the recruitment-relevant structure your feasibility team would otherwise build by hand.

  • Inclusion / exclusion criteria with normalization
  • Target population, geography, and site constraints
  • Visit burden, travel burden, and compensation parsing
  • Channel and policy risk detection
2

Recruitment feasibility modeling

Eligible population, conversion funnel, and enrollment timeline — with confidence bands rooted in primary sources, not vibes.

  • Eligible-population sizing with Census MOE bands
  • Prevalence ranges from CDC / NIH / specialty society literature
  • Geographic feasibility & competitive trial landscape (ClinicalTrials.gov)
  • Difficulty scoring (1–10) with risk and positive-factor drivers
3

Recruitment strategy generation

A full, protocol-aligned plan — not just ad copy. Channel mix, geographic rollout, enrollment pacing, and cost modeling.

  • Channel recommendations (Meta, Search, referral, site)
  • Geographic rollout and metro-expansion logic
  • Enrollment pacing model tied to your timeline
  • Budget scenarios with cost-per-PQL projections
4

IRB-ready materials package

Ad copy, landing-page content, prescreening, and patient-facing language — versioned, audit-logged, and aligned with your protocol from the start.

  • Compliant ad and outreach copy across channels
  • Landing-page and prescreener drafts
  • Consent-friendly patient-facing language
  • Versioned IRB submission bundle, exportable
Why sponsors trust the numbers

Methodology, confidence, and human-in-the-loop — by default

Every feasibility estimate is cited. Every AI output is editable. Every change is logged. Built to clear sponsor governance, not work around it.

MCited methodology

Every population number, conversion rate, and timeline traces back to a primary source visible to your team.

Sources: US Census ACS 2019–2023, CDC / NIH prevalence literature, Pew Research digital-reach surveys, ClinicalTrials.gov v2 API.

CConfidence bands

Point estimates ship with ± bands derived from Census margin of error, prevalence range across the literature, and channel performance variance.

Example: "23,400 eligible adults ± 8%" — not a single number you have to defend in committee.

HHuman-in-the-loop

Every AI-drafted material has an explicit review state. Sponsor reviewers edit, approve, and version before anything leaves Horizon.

Controls: editable outputs, approval gates, side-by-side version diffs, who-changed-what audit log.

IIRB-aligned by default

Drafted to clear common IRB objections (compensation phrasing, image restrictions, audience-sensitive language) before submission.

Output: versioned submission bundle, revision response packets, approval-state tracking per cycle.

AIndication-aware

Different logic for rare disease vs. common condition, pediatric vs. adult, oncology vs. cardiometabolic — not a one-size-fits-all funnel.

Coverage: 60+ indication prevalence profiles, age-banded digital-reach curves, peripartum-specific population logic.

SSponsor-grade auditability

Append-only event log on every state change. Pricing provenance, source logging, rollback support, confidence-score history.

Why it matters: survives sponsor governance review, external audit, and IRB inspection without rework.

Who Horizon AI is built for

Sponsor-side teams who own feasibility, study startup, and recruitment risk

Horizon AI replaces weeks of manual feasibility work and recruitment-vendor back-and-forth with a structured, defensible workflow. It sits inside your team — not your CRO's.

  • VP

    VP & Director, Clinical OperationsWants enrollment assumptions de-risked before site activation. Wants protocol viability questioned in hours, not after the first interim.

  • SSU

    Study Startup & Feasibility leadsReplace manual prevalence pulls, geographic feasibility decks, and IRB-material drafting cycles with one auditable workflow.

  • CO

    Clinical Operations & StrategyStandardize how recruitment is planned across programs. Reduce vendor variance. Make timelines and budgets defensible in committee.

Where Horizon AI fits

Upstream of your CROs, sites, and recruitment vendors

Horizon AI is the planning and IRB-readiness layer. It hands off cleanly to any execution partner — including CliniContact for execution, or your existing vendors.

Sponsor side — Horizon AI

Protocol & feasibility

Protocol ingestion, eligible-population modeling, recruitment strategy, IRB-ready materials with confidence bands and audit trail.

CRO / site network

Site selection & activation

Your existing CRO, site network, and study startup vendors continue to own site activation, contracting, and start-up logistics.

Recruitment vendors

Execution & enrollment

Recruitment vendors — CliniContact or any partner — execute against the validated strategy and IRB-approved materials Horizon produced.

Vendor-neutral by design. Horizon AI is the sponsor's strategy and IRB-readiness layer. It outputs to any execution partner. We don't lock you into a recruitment vendor, and the work product travels with you across studies and sponsors.

Most trials fail or delay because feasibility assumptions were wrong from the start.

Horizon AI fixes that at the protocol stage — with cited methodology, confidence scoring, and IRB-ready outputs your team can defend.