Uploaded July 2026 | Updated September 2026, 2 weeks ago
This demo shows how Mediforce turns a ClinicalTrials.gov NCT ID into reproducible synthetic SDTM datasets, with human review built into the workflow.
Created for CDISC AI Innovation 2026, the workflow retrieves the study record through the ClinicalTrials.gov API and uses AI to create a structured study configuration. A reviewer can inspect and adjust that configuration before the data generation begins.
The approved configuration is then processed using the open-source synthsdtm R package with a fixed seed, making each run reproducible. The generated datasets are checked with CDISC CORE, and the reviewer can examine the domains, findings, conformance results, and downloadable reports before approving the final output or requesting changes.
The video covers:
• Retrieving study metadata from an NCT ID
• Creating and reviewing a synthetic study configuration
• Generating reproducible SDTM datasets
• Running CDISC CORE conformance checks
• Reviewing datasets, findings, and validation reports
• Approving the output or requesting changes
Mediforce is an open-source platform for orchestrating AI agents in transparent, human-supervised workflows.
Learn more about Mediforce: mediforce.ai
Explore the project on GitHub: github.com/Appsilon/mediforce
#CDISC #ClinicalData #SyntheticData #PharmaAI #OpenSource
This demo shows how Mediforce turns a ClinicalTrials.gov NCT ID into reproducible synthetic SDTM datasets, with human review built into the workflow.
Created for CDISC AI Innovation 2026, the workflow retrieves the study record through the ClinicalTrials.gov API and uses AI to create a structured study configuration. A reviewer can inspect and adjust that configuration before the data generation begins.
The approved configuration is then processed using the open-source synthsdtm R package with a fixed seed, making each run reproducible. The generated datasets are checked with CDISC CORE, and the reviewer can examine the domains, findings, conformance results, and downloadable reports before approving the final output or requesting changes.
The video covers:
• Retrieving study metadata from an NCT ID
• Creating and reviewing a synthetic study configuration
• Generating reproducible SDTM datasets
• Running CDISC CORE conformance checks
• Reviewing datasets, findings, and validation reports
• Approving the output or requesting changes
Mediforce is an open-source platform for orchestrating AI agents in transparent, human-supervised workflows.
Learn more about Mediforce: mediforce.ai
Explore the project on GitHub: github.com/Appsilon/mediforce
#CDISC #ClinicalData #SyntheticData #PharmaAI #OpenSource
