AI Study Builder
The AI Study Builder helps you design and configure clinical studies directly in Ledidi, within GCP guidelines. Describe the study you are running and the builder drafts the structure: visits and visit windows, unplanned events, the forms attached to each of them, and participant-reported surveys. Your team reviews and edits everything before it goes live. Built for research teams and CRO study managers rather than programmers.
How it works
01
Describe the study
Tell the builder what you are running: the design, the intervention, the follow-up schedule and what you need to collect. Plain language, inside the platform.
02
Review the structure
Visits, visit windows, forms, unplanned events and participant surveys are drafted as editable Ledidi objects. Keep what works, change the rest.
03
Approve and publish
Nothing is published until you approve it. The study is then configured in Ledidi and ready for sites, coordinators and participants.
What the builder sets up
Each item is created as an editable object in your study, not a document you then have to build from.
Visits and visit windows
Screening and baseline, plus follow-ups scheduled relative to an intervention date, each with its own window.
Unplanned events
Adverse events, SAEs and other event-driven records that sit outside the visit schedule.
Forms on every visit and event
Fields, value ranges and validation, attached where they belong and editable throughout.
Participant surveys
Patient-reported outcomes configured on their own schedule and sent to participants directly.
The builder proposes. Your team decides.
AI in a regulated study only works if a qualified person can see what it did and say no. Every suggestion is reviewable, editable and attributable before it affects data collection.
Review before publish
No generated visit, form or survey becomes part of a live study until someone approves it.
Editable, not final
Everything the builder creates is an ordinary Ledidi object. Rename it, restructure it, delete it.
Full audit trail
Configuration changes are logged with user, timestamp and before/after state, the same as any other action in the platform.
Built for GCP
The build process is designed to fit GCP expectations and FDA 21 CFR Part 11 controls, not to work around them.