← Back
📁 MARK'S DOC
🔗 Linked Evidence
The_Pang_Formula_and_the_BEAST — p.8
📄 The Pang Formula and the BEAST | p.8
📄 The Pang Formula and the BEAST | p.8
📝 Extracted Text (OCR)
interconnected dataset with an underlying relational structure. In this framework, each document is a node, each reference a potential edge, and each date a timestamp that either verifies an administrative timeline or exposes a contradiction. Attempting to audit an investigative file of this scale with a highlighter and a legal pad guarantees that critical systemic connections will be missed. The administrative apparatus was counting on exactly that. "The Evidence Beast" (v3.6) is a purpose-built relational clinical database engineered specifically to ingest, relationalize, and instantly retrieve this multi-jurisdictional record, anchoring decentralized administrative actions back to the raw physical reality of the workplace. Technical Architecture — What the Code Does The Evidence Vault is built on a Python Flask backend with a SQLite database managed through SQLAIchemy ORM. Every piece of evidence in the file exists as a structured EvidenceCard record. Each card represents a document in the claim file—a hospital discharge note, an FOI page, a WCAT decision excerpt, an internal draft processing log, an audiogram. The cards are typed by evidence category (EMPLOYER, WORKSAFE, MEDICAL, UNION, LEGAL, INTERNAL, FOI, PERSONAL), each with its own colour code in the interface, so the provenance of every piece of evidence is visible at a glance. The Bidirectional Link Engine — The Antidote to Decoupling The most strategically significant feature of the Evidence Beast is its bidirectional link synchronization. This was built in direct response to the core mechanism of the Pang Formula: the administrative practice of keeping related documents separated so that a reviewer will not connect the threads. When the worker linked the Flora Pang draft processing notes (FOI p.24) to the final Clinical Opinion (Claim File p.305-306), the system automatically created the reverse link—ensuring that opening either document in the Vault immediately presents the other. The connection cannot be broken. The draft note and the sanitized final report are permanently joined. This function is the technical answer to administrative decoupling. The 1991 BC Ferry noise log cannot be read in isolation from the 2026 Clinical Opinion that cited it without dates. The internal draft note cannot be read in isolation from the final report that stripped its timestamps. Every time either document is opened, the system surfaces the other. This is precisely what the Board's data management workflow was designed to prevent. The Evidence Beast was designed to make prevention impossible. Document Ingestion and OCR Pipeline The system supports full document ingestion via PDF and image upload. When a document is imported, it is processed through an OCR pipeline using Tesseract and pdf2image, converting each page into a searchable text record stored alongside the original image.