Trial team solutions
Optimizing clinical trial setup through intelligent automation and cubeCDMS integration
EDC build
Process automation
Data management
Background
SevenTwoOne Inc. is a CRO with 74 employees providing one-stop solutions to pharmaceutical companies across consulting, outsourcing, and IT solutions.
EDC development and data management historically relied on time-consuming, manual workflows. The company faced operational bottlenecks across specification creation, test data generation, EDC build, and paper survey pre-printing.
To overcome these challenges, SevenTwoOne streamlined its development lifecycle by leveraging the unique strengths of AI, Robotic Process Automation (RPA), Microsoft Office, and cubeCDMS.
The following pain points were the main drivers for change:
High transcription volume in specification creationCreating database specifications from questionnaires required extensive manual transcription, a highly inefficient process with high risk of errors. |
Discover how SevenTwoOne, a CRO, uses AI, RPA, Microsoft Office, and cubeCDMS to transform EDC development, reducing manual effort across specification creation, test data generation, EDC builds, QC, and paper survey processing.
Background
SevenTwoOne Inc. is a CRO with 74 employees providing one-stop solutions to pharmaceutical companies across consulting, outsourcing, and IT solutions.
EDC development and data management historically relied on time-consuming, manual workflows. The company faced operational bottlenecks across specification creation, test data generation, EDC build, and paper survey pre-printing.
To overcome these challenges, SevenTwoOne streamlined its development lifecycle by leveraging the unique strengths of AI, Robotic Process Automation (RPA), Microsoft Office, and cubeCDMS.
The following pain points were the main drivers for change:
High transcription volume in specification creationCreating database specifications from questionnaires required extensive manual transcription, a highly inefficient process with high risk of errors. |
Increased effort and missing patterns in test dataGenerating edit check (EC) test data manually across growing numbers of EC rules increased workload and heightened the risk of missing or incorrect test conditions. |
Inconsistent quality control (QC) standardsInconsistent QC standards across reviewers made it easy for formatting and template errors to slip through human review. |
Heavy data entry workload in EDC buildManually configuring the EDC created high operational strain and was also an error-prone process. |
Pre-print burden in paper-based surveysPre-printing paper questionnaires by hand created a massive burden for data managers, consuming roughly 4,000 hours of manual labour for a 500-case, 8-volume study. |
Solution: Integrated AI, RPA & cubeCDMS
SevenTwoOne broke down the development lifecycle from project setup to release into six distinct stages, leveraging AI, RPA, Microsoft Office, and cubeCDMS to optimize these processes.
The result is a smoother, more efficient workflow:
AI-drafted database specifications
Using Copilot Studio, survey layouts are structured into JSON and split into modular processing units (Visit list, Page/CRF list, Item list), directing instructions to child agents to generate highly accurate specification drafts.
Multi-stage AI test data generation
AI agents use a multi-stage request process with human intervention to generate comprehensive test patterns covering complex edit check condition logic (AND/OR, boundary checks, multiple domains).
Standardized document QC
Dedicated AI agents process Word checklists into structured Excel issue logs focused strictly on formatting and notation inconsistencies.
Automated EDC build via RPA
Power Automate Desktop (RPA tool) reads embedded parameters in specifications to automatically populate fields, code lists, and properties directly in cubeCDMS builder.
Automated pre-print tools
Power Query and Excel VBA automate data extraction, template population, and file output, eliminating individual expertise dependencies.
Results
Implementing an integrated automation strategy delivered clear operational and quality benefits across SevenTwoOne's workflow:
70–80% reduction in effort for specification document creation.
60–80% reduction in workload for test data creation.
90% reduction in DB build time and 85% reduction in ECS build effort.
75% reduction in hours for paper survey preprint processing.
Improved QC Consistency: Standardized feedback across deliverables to eliminate subjective review criteria and ensure complete template compliance. |
A major achievement was building in-house capabilities to develop custom AI workflows and agents. By leveraging these tools to deliver early prototypes, SevenTwoOne secured faster client feedback, freeing up staff to focus on strategic discussions and validation prep.
As one of SevenTwoOne’s representatives noted, "Our main hurdle was figuring out how to sequence the AI's workload. Feeding it everything at once diluted the output, but by breaking the process into stages and giving targeted instructions to specific child agents, we got extremely accurate results."
Future outlook
SevenTwoOne plans to further standardize specification creation and expand RPA application across tasks. Their emphasis is to restrict AI processing to internal tenants and to include strict Human-in-the-Loop oversight to guarantee data security and quality control.
Discover how SevenTwoOne, a CRO, uses AI, RPA, Microsoft Office, and cubeCDMS to transform EDC development, reducing manual effort across specification creation, test data generation, EDC builds, QC, and paper survey processing.
