What problem does it solve? Compliance frameworks like SOC 2 and NIST 800-53 arrive as spreadsheets, CSVs, or PDFs with inconsistent column names and layouts, making it hard to turn raw requirement rows into structured data that downstream control-design workflows can consume. ## Core Features & Use Cases - Multi-format ingestion: Reads Excel (.xlsx), CSV, and structured text/PDF documents, mapping varied column headers (ID, Title, Description, Category, Evidence) to a normalized schema. - Keyword extraction: Strips stopwords and retains technical and compliance-domain terms, producing 5-10 keywords per requirement for matching against existing controls. - Automation classification: Labels each requirement as automated-evidence, manual-attestation, hybrid, or informational, with a confidence score that gates downstream action (proceed, confirm, or flag for review). - Use Case: Ingest a SOC 2 spreadsheet, normalize every row into structured JSON with keywords, and classify each requirement so the converting-frameworks-to-controls workflow knows which controls to design automatically and which need human attestation. ## Quick Start Parse the attached SOC 2 framework spreadsheet and classify each requirement by automation feasibility with confidence scores.
