Sign In

Introduction

Tracore is an API for parsing documents and extracting structured data. You define a schema that describes the data you want, upload a document, and Tracore returns clean, validated JSON.

Extraction is powered by AI and runs asynchronously. You submit a job and poll for results, or receive them via webhooks.

Core concepts

Workspaces

A workspace is a top-level container for your project. All schemas, documents, and runs belong to a workspace. Each workspace has a unique slug used in API URLs.

Schemas

A schema defines the structure of data you want to extract. Schemas use JSON Schema format and support versioning — every change creates a new immutable version, so you always know exactly which definition produced a given result.

Documents

Documents are the files you upload for processing. Tracore supports PDFs, images, and text files. Documents are scoped to a specific environment within a workspace.

Runs

A run represents a single extraction job. When you submit a document for extraction, Tracore creates a run that moves through statuses: pending, processing, and finally completed, failed, or validation_failed.

Environments

Each workspace has three environments: production, staging, and development. Documents and runs are scoped to an environment, while schemas are shared across all environments. This lets you test with sample documents in development before processing real data in production.

Webhooks

Webhooks notify your application when events happen — such as a document being received or an extraction completing. Webhooks are scoped to a specific schema and environment.

Next steps