MediLake

MediLake is an all-in-one Clinical Data Management System (CDMS) developed by CLUPEA, designed for pharmaceutical companies, biotechs, CROs, hospitals, and academic research institutions conducting clinical trials from Phase I through global regulatory submissions. The platform integrates EDC, CTMS, eTMF, IWRS, ePRO, eConsent, pharmacovigilance, and CDISC data services into a single cloud-based solution. Established in Korea in 2017, MediLake has processed over one million clinical metadata records and holds CDISC ODM Export/Import certification, with compliance covering FDA, EMA, GAMP5, ICH E6(R2), and 21 CFR Part 11 requirements.

A central feature of MediLake is its AI-powered eCRF auto-generation capability, which combines CDISC ODM certified technology with machine learning to reduce EDC build time by approximately 80%, from six weeks to one week. The platform is built natively on CDISC standards, supporting automatic SDTM, ADaM, and SEND dataset generation throughout the trial lifecycle.

AI eCRF Generation

• Protocol upload triggers automatic eCRF generation based on CDASH and SDTM standards
• AI produces end-to-end SDTM Specifications from uploaded CRF content
• Edit Check Validation is auto-configured as part of the generation process
• EDC build time reduced from six weeks to one week (80% reduction)

CDMS / EDC

• WYSIWYG CRF Designer with drag-and-drop editing and a library of 30+ widgets supporting grid and column responsive layouts
• Conditional logic, repeating groups, derivation rules, and edit check rules configurable without coding
• Real-time Dataviewer with visual Merge/Join across datasets, requiring no SQL knowledge
• Supports data monitoring and cross-domain exploration for data managers, CRAs, biostatisticians, and principal investigators
• AI Edit Check Validation using per-item deep learning models to auto-suggest range checks, cross-field validations, and consistency checks
• AI Chatbot interface accepts natural language input (e.g., "Add range validation for systolic BP") to generate edit checks instantly
• Auto-configures cross-domain validation, conditional range checks, and visit-sequential consistency rules without coding
• Open API Gateway using RESTful API and WebSocket for bidirectional integration with PV, Safety DB, CTMS, and BI systems
• OAuth 2.0 / JWT secured; supports AE/SAE auto-reporting, MedDRA transfer, and lab data transfer
• E2B(R3) ICSR auto-generation with regulatory authority transmission
• HL7 FHIR R4 EHR/EMR connectivity and CDISC ODM v2.0 metadata exchange
• Audit trail and multi-site support included

Clinical Trial Management System (CTMS)

• Interactive Gantt chart with critical-path analysis and automated dependency management
• Automated progress tracking across Startup, Enrollment, Treatment, and Close-out phases with KPI dashboards
• Role-Based Access Control (RBAC) for PI, CRA, CRC, and Biostatistician roles with approval workflow automation
• Site feasibility assessment, contract management, and IRB/EC approval timeline tracking
• Built-in real-time communication and issue tracker between sponsor and sites
• Bidirectional data integration with EDC, eTMF, IWRS, and Safety modules

Electronic Trial Master File (eTMF)

• Fully aligned with CDISC TMF Reference Model v3.3.1 (2023)
• 11-Zone, 58-Section, 283-Artifact hierarchy with automated classification and indexing
• ICH E6(R2) GCP and ISO 14155 code mapping for inspection-ready document management
• Multi-level metadata (Trial/Country/Site) with milestone-driven auto-filing
• 21 CFR Part 11 compliant e-signatures and full audit trails for global multi-site TMF completeness

Electronic Patient-Reported Outcomes (ePRO)

• No-code drag-and-drop WYSIWYG Form Builder supporting standard instruments including PRO-CTCAE, EQ-5D, and VAS
• Visual branching logic and skip pattern configuration
• Simultaneous Android and iOS native app deployment on live publish
• BYOD and provisioned device support with offline data capture and auto-sync
• Multilingual support and compliance alerts

Interactive Web Response System (IWRS)

• AI-powered stratified randomization with covariate-adaptive minimization
• Automated resupply alerts with blinding protection
• Real-time inventory tracking and drug supply optimization
• Demand forecasting for Investigational Medicinal Product (IMP) supply management

Electronic Informed Consent (eConsent)

• NLP engine-based consent form auto-generation with readability scoring
• Medical-specialized multilingual translation and patient comprehension assessment
• Remote signing with version control for protocol amendments
• Multimedia content (video and images) to support patient understanding

CDISC Data Services

• CDISC ODM certified for 6 cases (3 Import, 3 Export), enabling full CDISC-based data transformation
• MediLake AI trained on 40+ SEND productions auto-generates SEND datasets from nonclinical study reports (repeat-dose toxicity, carcinogenicity, genotoxicity, and other study types)
• SENDIG-compliant, covering DM, BW, CL, FW, LB, MA, MI, OM, TF, TS, and all required domains
• Converts clinical data from Rave Medidata, Oracle InForm, Veeva Vault, and other EDCs to CDISC SDTM standard datasets via AI Auto-Mapping Engine
• Automated domain detection, variable mapping, controlled terminology alignment, and codelist matching
• Produces submission-ready data packages for FDA, EMA, PMDA, and other global regulatory authorities
• ADaM analysis dataset generation and eCTD submission support

Real-World Data and Evidence (RWD/RWE) Platform

• Ingests EHRs, claims, registries, wearable IoT, and patient-generated health data at petabyte scale
• Distributed computing with columnar storage enabling sub-second queries across billions of records
• In-memory caching and parallel pipelines for linear scalability in cohort and longitudinal studies
• OMOP CDM and FHIR/HL7 mapping for clinical data interoperability
• Cohort analytics engine with automated regulatory reporting; HIPAA and GDPR compliant

Pharmacovigilance and Safety Management

• MedDRA-coded AE/SAE registration with causality assessment using Naranjo and WHO-UMC scales
• AI-powered safety signal detection using PRR (Proportional Reporting Ratio) and BCPNN (Bayesian Confidence Propagation