TrialOS
start now

Stop guessing when your trial will finish enrolling.

Enrollment challenges often emerge months before a study misses its targets. We use enrollment data to forecast study completion dates, estimate milestone probabilities, and compare intervention scenarios to improve outcomes.

simulate studylearn more
trialos — studies / ONCO-204
ONCO-204 · target N 200
Forecast: P80 Mar 12, 2027
on-time prob.
62%
target N = 200nowP80 · Mar 27
Adjustment Plan
  • Add 2 sites @ median rate−41 days
  • Boost top quintile +15%−22 days
  • Retire SITE-019 · redistribute−7 days
approve plan

TrialOS Core. the open foundation everything runs on.

The forecasting and scenario engine are free, open source, and yours to host. One command gives you realistic completion dates, site-level risk, and a simple API you can call from R, Python, or your own tools.

~/trial-ONCO-204init → forecast · 28s
$ pip install trialos
  ✓ resolved 12 packages in 1.2s

$ trialos forecast enrollments.csv --target-n 200
  → loaded 14 sites · 312 weekly observations
  → fitting site rates (poisson-gamma)............. ok
  → simulating 10,000 enrollment paths.............. ok

  forecast: ONCO-204
  ─────────────────────────────────
  median completion   2026-12-04
  P80 completion      2027-03-12
  on-time probability 62%
  bottleneck sites    SITE-019, SITE-007

  wrote → forecast.json · curve.png · sites.png
Out of the box
  • Monte-Carlo enrollment simulation
  • Site-level rate inference
  • What-if scenarios via simple API
  • Forecast JSON + plots, no account

Everything stays in your repo, under your control, versioned alongside your code.

If you want agentic infrastructure, automated re-forecasting, and adjustment plan insights, that's the hosted layer. But the engine is open, and it always will be.

the workspace where your trials get planned.

  1. 1

    Use your enrollment data.

    Import site-level enrollment counts and study metadata. No patient-level data required.

  2. 2

    Generate a forecast.

    TrialOS simulates enrollment trajectories to estimate completion dates, uncertainty ranges, and milestone probabilities.

  3. 3

    Evaluate interventions.

    Test scenarios such as adding sites, increasing enrollment rates, or reallocating resources. See the projected impact on timelines and success probability.

  4. 4

    Compare outcomes.

    Understand which actions are most likely to improve study performance before committing time and budget.

commonly asked questions.

Is this a CTMS?
No. TrialOS doesn't replace your CTMS or EDC. It sits beside them as an enrollment forecasting and planning layer.
Will this touch PHI?
No. TrialOS is built to work on aggregate site-level counts only.
Who is this for?
Sponsors, CROs, and academic units who care about enrollment risk and want something better than spreadsheet trendlines.
Do we have to use the hosted version?
No. You can run the core library entirely on-prem and never talk to our servers.

Know sooner. Enrollment delays rarely appear suddenly.

We help teams identify risk earlier, quantify its impact, and evaluate responses while there is still time to act.