Privacy-first document refinement. Browser-side entity detection, PII anonymisation, and intelligent polishing for professionals who handle sensitive information.
meaning preserved · tone refined · nothing lost
Powered by Flock-7 v1.7.0 · Privacy by Design · Full Audit Trail
Context-aware correction that goes beyond standard spell-checkers. Handles professional terminology, industry jargon, and regulatory language across all five supported sectors. Catches homophone errors, subject-verb disagreements, and tense inconsistencies that automated tools typically miss.
Restructures convoluted sentences and simplifies complex passages while preserving technical accuracy and original meaning. Breaks down lengthy paragraphs, eliminates redundancy, and improves logical flow — particularly effective on dense regulatory correspondence and clinical documentation where readability directly affects compliance.
Matches the appropriate register for your audience and sector. Professional, formal, empathetic, or neutral — consistently applied throughout the document. Sector-specific calibration ensures insurance correspondence reads differently from medical referral letters, with vocabulary and formality tuned to industry expectations.
Clinical Entity Recognition engine trained on international standards including SNOMED CT, ICD-10, dm+d, and Lloyd's market terminology. Pattern recognition handles jurisdiction-specific identifiers across European regulatory frameworks. Detects names, dates of birth, NHS numbers, policy references, financial identifiers, addresses, and medical diagnoses — all processed locally in the browser before any cloud transmission.
Detected entities are replaced with coded placeholders before transmission using the Flock-7 anonymisation layer. The mapping between original values and placeholders is held exclusively in browser memory — never transmitted, stored externally, or logged. After polishing, original entities are restored locally with integrity verification confirming every placeholder was correctly reinserted.
Every processed document generates a downloadable Trust Certificate containing a unique SHA-256 cryptographic hash, processing metadata, entity counts, sector classification, and verification status. Provides an auditable compliance artefact that can be presented to regulators, DPOs, Caldicott Guardians, or internal audit teams as evidence of privacy-compliant processing.
Complete processing chain recorded for every document — entity detection, anonymisation, cloud transmission, polishing, restoration, and verification — all traceable via the Trust ID. The real-time systems panel displays each stage with status confirmation. No document content is stored; only cryptographic hashes and processing metadata are retained for audit purposes.
Full polishing and entity detection across major European languages including English, German, French, Spanish, Italian, Dutch, and Portuguese. Same privacy guarantees, same anonymisation pipeline, same audit trail regardless of language. CER entity patterns support international date formats, European address structures, and jurisdiction-specific identifier formats.
Zero-retention architecture from the ground up. Documents are anonymised in the browser before anything touches the cloud — no server-side storage, no logging, no third-party data sharing. The processing pipeline is engineered for UK GDPR, EU AI Act transparency requirements, FCA data handling expectations, and NHS Caldicott Principles from the architecture level, not as a compliance afterthought.
Comprehensive entity detection built on a cautious-by-default philosophy — if it could be sensitive, it gets protected. Validated across English, German, French, and Spanish documents with support for all major European languages. Detection models trained on regulated sector terminology spanning medical, insurance, legal, and financial vocabularies.
People, companies, organisations
Email, phone, fax numbers
Street, city, postcode, country
Policy, claim, account, reference
Amounts, account numbers, IBAN
DOB, incident, deadline dates
Diagnoses, treatments, conditions
Case numbers, court references
Input your document text
CER scans for entities (browser-side)
Flock-7 replaces PII with placeholders
AI refines the anonymised text
Original entities reinserted locally
DocPolish serves professionals who can't afford to expose client data to cloud AI.
Claims correspondence, policy wordings, underwriting notes, bordereaux submissions, loss adjuster reports, cover notes
Client letters, case summaries, contract reviews, witness statements, advice notes, court submissions
Advisory reports, compliance documentation, client communications, regulatory submissions, audit correspondence
Patient letters, referral notes, clinical summaries, discharge letters, GP correspondence, research abstracts
Research papers, grant applications, student feedback, ethics submissions, peer review responses, abstracts
From quick emails to detailed reports—DocPolish handles them all.
Professional tier ensures your sensitive data never leaves your browser. Verified by cryptographic audit trail.
Entity Detection
Browser-side rules engineAnonymisation
Placeholder substitutionTo Cloud
Guaranteed isolationHash Chain
Verifiable audit trailSee our transparent pricing or start polishing immediately.