Built by Metorial, the integration platform for agentic AI.
Upload a document to Affinda for AI-powered parsing and data extraction. Supports uploading via a publicly accessible URL. The document will be processed according to the workspace or collection configuration, extracting structured data such as resume fields, invoice line items, or other document-specific data points. Set **wait** to `true` to receive parsed results immediately, or `false` to get the document identifier and poll later.
Update multiple document annotations in a single request. Use this to programmatically correct or confirm extracted data points. Each update specifies an annotation ID and the new values to set.
Permanently delete a document from Affinda. This removes the document and all its extracted data from the database. This action cannot be undone.
Get a compatibility score between a specific resume and a specific job description. Returns an overall match score (0 to 1) along with category-level breakdowns for skills, experience, education, and more. Both the resume and job description must already be uploaded and parsed in Affinda.
Search through parsed resumes or job descriptions using Affinda's Search & Match algorithm. Match resumes against job descriptions or vice versa, or search using custom criteria like job titles, skills, experience, education, and location. Returns a ranked shortlist with matching scores for each category.
Retrieve all annotations (extracted data points) for a specific document. Each annotation represents a field extracted by the AI model, such as a name, date, amount, or address.
List and search documents in Affinda. Filter by workspace, collection, processing state, tags, or search by filename. Supports pagination and sorting. Use this to browse uploaded documents or find specific ones.
Generate a redacted version of a parsed resume document. Select which categories of personally identifiable information (PII) to redact, including personal details (name, address, phone, email), work details (company names), education details (university names), headshots, referees, locations, dates, gender, and PDF metadata. Returns a base64-encoded PDF with the selected fields redacted. The original document is not modified.
List all workspaces within an organization. Workspaces are logical project containers where documents are processed. Optionally filter by name.
List all organizations accessible with the current API key. Organizations are the top-level containers that hold workspaces, document types, and team members. Use this to discover organization identifiers needed for workspace and document type operations.
Retrieve a specific document and its extracted data from Affinda. Returns document metadata, processing state, and the structured data extracted by the AI model. Use this to check parsing status or fetch results for a document uploaded with `wait: false`.
Create a new workspace within an organization. A workspace is a logical container for document processing that can have its own configuration, document types, and access permissions.
Permanently delete a workspace and all its contents. This removes the workspace, its collections, and all documents within it. This action cannot be undone.
List available document types (extractors) configured in your Affinda account. Document types define the AI model configuration for specific document kinds such as resumes, invoices, bank statements, passports, etc. Use this to discover available document type identifiers for uploading documents.