Everything you need
to build stronger cases.

MedCase accepts PDF documents up to 2 GB, including scanned and handwritten documents. The system automatically detects whether pages need OCR and applies it where needed. You can also force OCR on all pages for PDFs with unreliable embedded text.

MedCase uses Claude Sonnet 4 by Anthropic for its protocol compliance analysis, running seven parallel analyses simultaneously to maximise coverage. Every finding includes direct evidence citations with page references and specific protocol section references, so qualified reviewers can verify each finding against the original document. MedCase is designed to flag potential protocol deviations for professional review. It supports qualified legal and medical professionals rather than replacing them.

Yes. MedCase AI includes intelligent OCR detection that automatically identifies scanned and handwritten pages. When OCR is needed, each page is rendered at 300 DPI and processed through an enterprise OCR engine. You can also force OCR on all pages for PDFs with unreliable embedded text layers. The system tracks which pages required OCR so you can assess extraction quality.

MedCase AI supports analysis against any clinical protocol that is uploaded to the platform. This includes NICE guidelines, Royal College guidelines, local NHS trust protocols, and other clinical standards. Protocols are chunked and embedded as vectors, enabling semantic search, so the system finds relevant protocol sections even when the medical record uses different terminology.

Every case includes an AI chat interface powered by Retrieval-Augmented Generation (RAG). You can ask natural language questions about any medical record and receive answers grounded in the actual document with page-level citations. The chat remembers previous messages for follow-up questions, is isolated per case, and the history is saved across sessions.

Yes. MedCase AI is built for teams. Every account is organised around organisations with unlimited team members on all plans. Organisation owners can control case visibility, either allowing all members to see all cases or restricting visibility so members only see cases they created.

MedCase accepts PDF documents up to 2 GB per upload, and supports cases with multiple uploads. There is no fixed page-count limit. Typical clinical negligence cases run from a few hundred pages to several thousand pages, and larger record sets are processed with progress tracking.

MedCase is designed to produce grounded responses with source citations to reduce unsupported answers. Findings and chat answers cite the page of the underlying medical record and, where relevant, the specific protocol section. All outputs are intended as an analytical aid for qualified professional review, so reviewers can verify each citation against the original document. MedCase does not make absolute claims of "no hallucinations".

Yes. MedCase is an analytical aid for qualified legal and medical professionals. It does not constitute medical advice, a diagnosis, a legal opinion, or a determination of negligence. All outputs should be reviewed and verified by a qualified solicitor and/or medical professional before use in any legal matter.

All data is encrypted with AES-256-GCM at rest with per-record nonces and tamper detection. A triple-layer PII sanitisation system removes patient-identifiable information before any AI processing. MedCase is built to support UK GDPR requirements, with comprehensive audit logging and European hosting.

No. All data is stored and processed within Europe. MedCase AI is registered with the UK Information Commissioner's Office (ICO) and does not transfer patient data outside European jurisdiction. PII sanitisation ensures that no patient-identifiable information is sent to any external AI service.

Before any medical record text reaches an AI model, it passes through three independent detection systems: Microsoft Presidio (an enterprise PII engine), spaCy named entity recognition, and 30+ custom regex patterns designed specifically for UK medical records. This layered approach catches NHS numbers, names, addresses, dates of birth, GMC numbers, postcodes, and dozens of other identifier types. The result is comprehensive PII removal with minimal risk of leakage.

The default retention period is 7 years to align with typical clinical negligence limitation periods. Retention is configurable per organisation, and complete case deletion is supported on request. End-of-contract handling is described in our DPA, available on request.

Yes. A Data Processing Agreement is available on all plans. Procurement teams can request a DPA, DPIA support, and our subprocessor list via the trust centre or hello@medcase.ai.

Yes. MedCase is built to support UK GDPR requirements, including PII sanitisation, audit logging, and European hosting. We process medical records under Article 6(1)(f) legitimate interest and Article 9(2)(f) for legal claims involving special category health data. We implement data minimisation through PII sanitisation, maintain configurable data retention policies (default 7 years), support the right to erasure with complete case deletion, and maintain comprehensive audit trails. We are registered with the ICO and all data stays within European jurisdiction.