CASE STUDY
TDCT Validator
A Python validation engine executing 20+ automated protocol validation rules using YAML-driven logic — published as a Zenodo preprint.
20+
Automated validation rules encoded
100%
YAML-driven, no hardcoded logic
2025
Published as Zenodo preprint
The Problem
Clinical trial protocols often contain design inconsistencies that go undetected until late in the review process, leading to costly delays, regulatory pushback, and rework after a trial has already begun.
The Solution
I built a Python validation engine that encodes clinical trial design rules as YAML configuration, allowing over 20 automated checks to run against any protocol document and flag inconsistencies before submission.
Rule Engine
20+ automated validation rules encoded as YAML-driven validation logic.
Schema Enforcement
Validates protocol structure against clinical trial design standards before submission.
Flaw Detection
Flags inconsistencies and design flaws before clinical trial initiation, reducing costly rework.
Structured Reporting
Generates a structured validation report summarising every rule outcome for reviewers.
How It Works
Protocol Ingested
A clinical trial protocol document is parsed into structured data fields.
YAML Rules Loaded
20+ validation rules are loaded from YAML-driven configuration files.
Automated Checks Run
Each rule is executed against the protocol to detect design inconsistencies.
Report Generated
A structured report flags every failed rule with explanation and severity.
Reviewer Sign-off
Reviewers act on flagged issues before the trial protocol is finalised.
Outcome
TDCT Validator gives clinical trial teams an automated first line of defence against protocol design flaws, reducing manual review time and catching issues before they become expensive problems. The work was published as a Zenodo preprint in 2025.