Forge Phase‑Ready|
Transform plain‑English protocols into simulated trial designs, synthetic cohorts, power curves, and CONSORT‑style flows — instantly.
For research and educational use only • Not clinical decision support
0+
Simulated scenarios
0%
Simulation coverage
0ms
Avg. response time
TrialForge Simulator
AI‑driven clinical trial engine
AI Protocol Analyzer
Powered by GPT-4 & Clinical Expertise
Describe your trial protocol in natural language. Our AI will parse parameters, run simulations, and provide actionable insights.
Trusted by clinical & research teams
Used by trial designers, biostatisticians, and innovation teams at leading academic and industry organisations.
Built for modern clinical research
Everything you need to design, pressure‑test, and communicate clinical trials with AI precision and transparent assumptions.
GPT‑4 Protocol Intelligence
Parse complex natural‑language protocols into structured design parameters in seconds.
- Natural language protocol parsing
- Arm, endpoint & schedule extraction
- Context‑aware suggestions
Monte Carlo Simulation Engine
Run thousands of simulated trials under different assumptions to de‑risk design choices.
- Power & sample‑size exploration
- Dropout and adherence scenarios
- Sensitivity analysis
Interactive Trial Dashboard
Explore power, cost, and risk metrics in one interactive view made for review meetings.
- Real‑time parameter tweaks
- Exportable reports & CONSORT views
- Collaboration‑ready summaries
How TrialForge AI Works
Step 1 / 3
Describe your trial
Start with plain English. Paste a protocol synopsis—arms, endpoints, follow-up, eligibility. TrialForge AI extracts the structure for you.
Why start here?
Your protocol already contains the seeds of every design decision. Instead of building a spreadsheet of parameters, TrialForge AI reads your own words and extracts what matters.
Natural language input
No rigid templates. You describe the trial; we extract the structure.
Works with existing docs
Paste from your protocol synopsis, SAP draft, or study notes—any format.
Refine iteratively
Change endpoints or follow-up timing without starting over.
✨ Processed transiently by default. Save/export only if you choose. For research and educational use only.
Step 2 / 3
AI extracts design parameters
In seconds, TrialForge parses your text into structured parameters—arms, endpoints, recruitment curves, dropout assumptions and more. Review, edit, or accept smart defaults.
Your protocol text
"Phase 2 RCT, 180 participants, HbA1c endpoint, 24-week follow-up…"
AI parsing engine
GPT-4 + domain logic extracts endpoints, arms, timing, and assumptions.
Structured parameters
Arms, endpoints, N, timing, dropouts—ready to simulate.
Your protocol text
"Phase 2 RCT, 180 participants, HbA1c endpoint, 24-week follow-up…"
AI parsing engine
GPT-4 + domain logic extracts endpoints, arms, timing, and assumptions.
Structured parameters
Arms, endpoints, N, timing, dropouts—ready to simulate.
What gets extracted
The AI reads your protocol like a statistician would—looking for design intent, not just keywords. Where details are ambiguous, it proposes clinically plausible defaults you can edit.
Typical extractions
- Trial phase & design
- Primary endpoint
- Secondary endpoints
- Study arms & ratio
- Target sample size
- Follow-up schedule
- Recruitment window
- Dropout assumptions
You stay in control
Every extracted parameter is editable. Change a default and downstream outputs (power, cost, risk) update immediately.
Study Design
Phase
Phase 2
Design
RCT (DB)
Arms
1:1 (200)
Primary endpoint
HbA1c
Timing & Recruitment
Enrolment period
12 months
Follow-up duration
24 weeks
Expected dropouts
~8–10%
Effect Size Assumption
Primary endpoint (HbA1c): 0.8% ± 1.5% (SD)
Example default based on comparable trial baselines.
Smart defaults
Suggests realistic assumptions to get you to a first simulation faster.
Transparent assumptions
Shows what was inferred vs explicitly stated, so reviews stay defensible.
Instant feedback
Edit a parameter and see simulated outputs update in real time.
Step 3 / 3
Run simulations & explore outcomes
Execute thousands of virtual trials. Stress‑test your assumptions. Visualize power, probability of success, recruitment risk, and timeline feasibility—all in one interactive dashboard.
Why Monte Carlo?
Classical sample size formulas assume tidy, perfect worlds. Real trials are messy: recruitment is lumpy, dropouts cluster, effect sizes vary. Monte Carlo simulation lets us run 10,000 virtual trials under realistic, noisy conditions—and see the full range of what could happen.
Explore any scenario
- →
Vary sample size
What if you recruit 150 instead of 180? See power drop in real time.
- →
Test recruitment curves
Optimistic site ramp‑up vs. conservative. How do timelines shift?
- →
Stress‑test assumptions
What if effect size is 20% smaller? Dropout rate doubles?
- →
Evaluate go/no‑go rules
Define decision thresholds now. See long‑run false positive & negative rates.
🎯 Why this matters
Most underpowered trials fail because teams relied on classical formulas that ignored real‑world noise. Simulation-driven design catches these risks before FPFV and saves millions in wasted trials.
Simulation dashboard
10,000 Monte Carlo runs | Primary endpoint power
Power vs. Sample Size
N = 120 → 240
Probability of Success (across scenarios)
Base case
87%
Conservative
62%
Optimistic
94%
Expected DB lock
Month 28
Recruitment at risk
12%
Instant recalculation
Change a parameter—any parameter—and watch 10,000 simulations re‑run in seconds. No waiting for lengthy computational cycles.
Scenario comparison
Run "base case" vs. "conservative" vs. "optimistic" side‑by‑side. Quantify the range of realistic outcomes so you can plan for contingencies.
Export & share
Generate CONSORT‑aligned reports, PDF dashboards, or JSON exports for your statistical team. All outputs are publication‑ready and fully reproducible.
Trial simulation
Phase 2 HbA1c efficacy study
Statistical Power
Target: 80% power achieved at N=170
Prob. of Success (primary endpoint)
Range: 78% – 94% (95% CI)
Database lock timeline
10th %ile
Mo 24
Median
Mo 28
90th %ile
Mo 34
Risk dashboard
Ready to simulate?
In three steps—describe, extract, simulate—you'll have a complete, defensible picture of your trial's operating characteristics. No spreadsheets. No guesswork. Just clarity.
Product
Features
Everything you need to design smarter trials
Purpose-built tools for researchers, statisticians, and programme teams. From protocol parsing to Monte Carlo simulation to publication-ready reports.
Natural language input
Paste or type your protocol synopsis in plain English. No rigid forms, no complex parameter trees. We parse the text.
AI parameter extraction
GPT-4 powered model learns your intent. Extracts arms, endpoints, timing, and assumptions with clinical expertise built in.
10,000 simulation runs
Execute thousands of virtual trials. Explore power, probability of success, recruitment risk, and timeline feasibility under realistic conditions.
Live interactive dashboard
Change a parameter. Watch simulations recalculate in real time. Explore power curves, probability distributions, and timeline ranges instantly.
Multi-scenario testing
Run base case, conservative, and optimistic scenarios side-by-side. Quantify the range of plausible outcomes for governance and planning.
Export & collaborate
Generate CONSORT-aligned reports, PDF dashboards, or JSON exports. Share with your statistical team, governance, or sponsors.
Built for real trial design
Every feature is grounded in regulatory precedent and clinical practice. We don't build flashy features—we build tools that change how you make decisions.
- ✓
Operating characteristics
Power, type I/II error rates, probability of success—computed across thousands of virtual trials.
- ✓
Recruitment & timeline forecasting
Stochastic recruitment curves show range of database-lock dates under realistic site ramp-up patterns.
- ✓
Sensitivity analysis
Vary effect size, dropout, or recruitment speed. See how conclusions hold under alternative assumptions.
- ✓
Adaptive trial design
Platform trial support, interim analysis rules, and dynamic arm allocation scenarios.
Example: Parameter sensitivity
Instant recalculation across all scenarios. No spreadsheets. No manual re-computation.
🔒 Privacy-first
No sensitive patient data required. Protocols are processed ephemerally and never stored without your consent.
📊 Reproducible
Every simulation is logged. Download full methodology and seed values for peer review and regulatory submission.
⚡ No dependencies
Works standalone. No need for SAS, R, or statistical software licenses. Pure web-native computation.
Solutions
Built for every stage of trial design
Whether you're in early concept, pre-FPFV planning, or managing adaptive protocols, TrialForge AI meets you where you are.
Early-stage sponsors & biotech
You have a hypothesis and a small dataset. Need to build a credible, defensible proof-of-concept protocol before raising capital or partnering.
How TrialForge helps
- ›Rapid go/no-go decision framework before protocol lock
- ›Quantify probability of success under real-world assumptions
- ›Create compelling investor/partner decks with simulation outputs
- ›Iterate protocol design in hours, not weeks
Academic researchers & CROs
You're designing multi-site, potentially platform trials. Need robust, transparent statistical reasoning that holds up to peer review and regulatory scrutiny.
How TrialForge helps
- ›Simulation-based justification for sample size & operating characteristics
- ›Platform trial support with multi-arm error control
- ›Export publication-ready methodology & appendices
- ›Collaborative dashboards for multi-site governance
Regulatory & statistical teams
You review protocols, negotiate with sponsors, and need to defend statistical approaches to regulators. Every assumption must be justified and reproducible.
How TrialForge helps
- ›Transparent, auditable assumptions behind sample size & power
- ›Full reproducibility: download seeds, code, and methodology
- ›Identify and de-risk overly optimistic assumptions early
- ›Facilitate governance conversations with clear scenario analysis
Programme & project managers
You own timelines, budgets, and risk. Need clear, data-driven forecasts of database-lock dates and feasibility under multiple scenarios.
How TrialForge helps
- ›Realistic timeline forecasting based on stochastic recruitment
- ›Scenario-based contingency planning (best, likely, worst case)
- ›Identify critical recruitment & dropout risks before FPFV
- ›Executive-friendly dashboards for steering committee & PMB
Feature comparison by role
| Feature | Sponsors | Academic | Regulatory | Programme |
|---|---|---|---|---|
| Natural language protocol input | ✓ | ✓ | ✓ | ✓ |
| AI parameter extraction | ✓ | ✓ | ✓ | ✓ |
| Monte Carlo simulation | ✓ | ✓ | ✓ | ✓ |
| Interactive dashboard | ✓ | ✓ | ✓ | ✓ |
| Multi-scenario analysis | ✓ | ✓ | ✓ | ✓ |
| Platform trial support | ✓ | ✓ | ✓ | — |
| Recruitment forecasting | ✓ | ✓ | — | ✓ |
| Full audit trail & reproducibility | — | ✓ | ✓ | — |
| Team collaboration & sharing | ✓ | ✓ | ✓ | ✓ |
| PDF/CONSORT export | ✓ | ✓ | ✓ | ✓ |
Which solution fits your team?
All roles benefit from the same core platform—designed to adapt to your workflow, not the other way around.
Pricing
Transparent pricing, no hidden fees
Simple, predictable plans designed for researchers at every stage. Pay for what you use. Cancel anytime.
Starter
Perfect for exploring simulation. Try the platform.
$29/mo
or $279/year (save $69)
- ✓Up to 3 concurrent simulations
- ✓10,000 Monte Carlo runs per simulation
- ✓Basic dashboard & export (PDF)
- ✓Email support (48hr response)
- —Advanced scenario analysis
- —Team collaboration
- —Priority support
Professional
Ideal for active trial design teams & sponsors.
$99/mo
or $949/year (save $239)
- ✓Unlimited concurrent simulations
- ✓50,000 Monte Carlo runs per simulation
- ✓Advanced dashboard, multi-scenario comparison
- ✓PDF, JSON, & CONSORT export
- ✓Team collaboration (up to 5 users)
- ✓Priority email support (24hr response)
- —Dedicated account manager
14 days free, no credit card required
Enterprise
For large CROs, academic consortia, and global programmes.
Custom pricing
Volume discounts & flexible contracts
- ✓Everything in Professional
- ✓Unlimited team members
- ✓Custom simulation runs & optimization
- ✓API access & custom integrations
- ✓Dedicated account manager
- ✓Priority 24/7 support & training
- ✓SLA guarantee & on-site options
Feature comparison
| Feature | Starter | Professional | Enterprise |
|---|---|---|---|
| Concurrent simulations | 3 | Unlimited | Unlimited |
| MC runs per simulation | 10,000 | 50,000 | Custom |
| Dashboard & export | Basic | Advanced | Advanced+ |
| Team users | 1 | Up to 5 | Unlimited |
| API access | — | — | ✓ |
| Custom integrations | — | — | ✓ |
| Email support | 48 hrs | 24 hrs | 24/7 |
| Dedicated manager | — | — | ✓ |
| SLA guarantee | — | — | ✓ |
Frequently asked questions
Can I change plans anytime?
Yes. Upgrade or downgrade at any time. Changes take effect at your next billing cycle.
What payment methods do you accept?
We accept all major credit cards, wire transfer, and purchase orders for Enterprise accounts.
Is there a free trial?
Yes. Professional plan includes a 14-day free trial with full access. No credit card required to start.
What happens to my data if I cancel?
Your data is yours. You can export all simulations and reports in PDF, JSON, or CSV before your account closes. We retain no data after cancellation.
Do you offer academic or non-profit discounts?
Yes. Reach out to our team at samoadeyemi@yahoo.co.uk and mention your institution. We offer 30-50% discounts for qualifying research.
What about HIPAA, GDPR, and compliance?
TrialForge AI is GDPR-compliant and does not require or process protected health information (PHI). Enterprise plans can discuss additional security requirements.
Ready to transform your trial design?
Start with Starter to explore. Upgrade to Professional when you're ready to scale. Enterprise available for large programmes.
Watch a Phase 2 HbA1c trial come to life
Follow a real-world style Phase 2 Type 2 diabetes protocol as TrialForge transforms it into a fully powered, risk-aware trial design in under 5 minutes.
Scenario: A biotech company needs a defensible Phase 2 HbA1c protocol before investor and governance meetings.
Step 1 · Paste protocol
Drop in the free-text protocol description. TrialForge extracts all design parameters in seconds.
Step 2 · Validate design
Quickly review phase, endpoints, arms, and timelines before running simulations.
Step 3 · Run 10,000+ simulations
Explore power curves, risk, and timelines across multiple design scenarios.
Sponsor protocol (free-text)
Phase 2 randomized, double-blind, placebo-controlled trial in adults with Type 2 diabetes. Arm 1: Agent XYZ 100mg daily (N ≈ 100) Arm 2: Placebo (N ≈ 100) Primary endpoint: Change in HbA1c from baseline to Week 24. Secondary endpoints: Fasting glucose, weight, lipids, safety labs. Visits: Weeks 0, 2, 4, 8, 12, 16, 20, 24. Duration: 24 weeks active + 4-week safety follow-up. Expected effect: 0.8% ± 1.5% HbA1c reduction (SD = 1.8%).
Structured protocol parameters
Extracted in ~5 secondsPhase
Phase 2 proof-of-concept
Indication
Type 2 diabetes
Design
Randomized, double-blind, placebo-controlled
Allocation
1:1 (Agent XYZ vs. Placebo)
Primary Endpoint
Change in HbA1c at Week 24
Secondary Endpoints
FG, weight, lipids, safety labs
Duration
24 weeks active + 4-week follow-up
Planned N
200 randomized
Evaluable N
~180 (10% dropout)
Effect Size
Cohen's d = 0.44
Target Power
87% (95% CI: 82–92%)
Simulation progress
8,432 / 10,000 runs
Running base scenario (~3 seconds remaining)
Statistical power
87%
95% CI: 82–92%
Phase 2 HbA1c base scenario (d = 0.44, N = 200, ~180 evaluable)
Scenario presets
- Phase 2 HbA1c (Base): d = 0.44, 87% power, N = 200 (≈180 evaluable)
- Conservative (d = 0.35): Lower effect size, ≈72% power at N ≈ 220
- Optimistic (d = 0.55): Higher effect size, ≈95% power at N ≈ 160
Power vs. sample size
Base case achieves 87% power at 200 randomized participants with an assumed 10% dropout, yielding ~180 evaluable participants.
Database lock timeline
Shows realistic lock dates under optimistic, median, and conservative operational assumptions.
Risk assessment
Borderline power if dropout exceeds 15%.
Recruitment sensitive to HbA1c thresholds and site activation speed.
Cost concentration around later visits and lab intensiveness.
Endpoints and follow-up duration aligned with standard diabetes PoC trials.
All risk views stay in the design and planning domain only—no real patient data is processed.
What you walk away with
Protocol-ready outputs
Download structured design, visit schedule, and key assumptions as AI-assisted protocol text.
Power and sample size appendix
Attach simulation plots and sensitivity analyses directly to your SAP or protocol appendix.
Risk and mitigation summary
Generate a concise risk narrative for governance, DMC, or investor review.
Export simulation-backed design justifications directly into your protocol, SAP, or governance packs—without touching PHI.
What just happened
A 5-minute, simulation-backed tour of your Phase 2 HbA1c protocol.
Start from a free-text protocol, and end with powered scenarios, risk views, and exportable documentation ready for reviewers.
Use with your own trial
Every design is unique. Paste your protocol and see how TrialForge reshapes power, timelines, and risk for your specific programme.
Free trial includes all features for 14 days.
Designed for trial designers, biostatisticians, and innovation teams
Use TrialForge AI to explore design options before first‑patient‑in, reduce amendment risk, and align stakeholders around the same synthetic evidence base.
Demo Protocol Used Throughout This Page
Phase
Phase 2
Design
RCT, 1:1 allocation
Endpoint
HbA1c change
Duration
24 weeks active
Planned N
200 randomized
Evaluable N
~180 (10% dropout)
Effect Size
Cohen's d = 0.44
Power
87% (95% CI: 82–92%)
Arms
1:1 (200)
Note: This illustrative Phase 2 HbA1c protocol is used consistently throughout TrialForge for demonstration. All parameters should be reviewed by a qualified biostatistician and approved by regulatory teams before use in any actual clinical trial.
Typical workflow
- •
Describe your trial in plain English
Paste or type a protocol synopsis, including arms, endpoints, follow‑up, and key eligibility criteria.
- •
AI extracts design parameters
TrialForge parses the text into structured parameters that can drive simulations and dashboards.
- •
Run Monte Carlo simulations
Explore power, sample size, dropout, and feasibility scenarios across thousands of virtual trials.
- •
Review, export, and share
Use dashboards, CONSORT‑style flows, and exportable reports to support internal review and iteration.
"TrialForge AI has become our sandbox for trial design — we can explore scenarios in hours instead of weeks and walk into review meetings with ready‑made visuals."
Senior clinical researcher, academic medical center