Introduction

Talonic is an agentic data platform that transforms unstructured documents into structured, validated data with per-cell provenance and AI reasoning traces. Whether you are processing invoices, contracts, purchase orders, or any other document type, Talonic automatically discovers every data point and maps it to a unified knowledge graph. The platform handles the entire pipeline from OCR and classification through structured output delivery, so you can focus on defining what data you need rather than building extraction logic.

Supported Formats: 25+ file types. Quick Jobs: 4 phases. Instant Matches: ~30% of cells (free). The platform ingests PDF, DOCX, XLSX, images (PNG, JPG), HTML, JSON, CSV, email formats (EML, MSG), and ZIP archives. ZIP files are unpacked automatically and each contained document is processed individually through the same pipeline.

Under the hood, Talonic ingests documents of any type, discovers every data point inside them, builds a knowledge graph of canonical fields, and runs AI extraction to fill structured output schemas. Every cell in the output carries provenance metadata — which pipeline phase filled it, the confidence score, and an AI reasoning trace linking back to the source document. The knowledge graph grows with every document you process — fields are clustered semantically, promoted through maturity levels, and enriched with master extraction instructions. This accumulated intelligence means extraction accuracy and speed improve over time without any manual tuning.

The platform follows a single flow in three movements: ingest, structure, review. You ingest on the Documents page, by upload or through connectors behind its Sources header button. You structure by composing a Spec in the Spec composer (/specs) and running it as a Pipeline, visible under Pipelines (/pipelines); for a quick one-off you can run a Job directly via Extract (Cmd+J). You review on Products, the review hub: its Status column shows which data products need review, each product opens in Review Mode, and the Decision log tab records every decision. Approved data then ships through delivery bindings. Most teams start by uploading a small sample of documents, composing a Spec, and running a preview. As the knowledge graph matures, an increasing share of cells are resolved instantly through graph lookup rather than AI extraction.

• 25+ file formats — PDF, DOCX, XLSX, images, HTML, JSON, CSV, email formats (EML, MSG), and ZIP archives.
• 4-phase quick Jobs — resolve from the knowledge graph, extract with AI directed by the Planner, re-resolve, then transform and validate.
• ~30% instant matches — cells filled from graph lookup are free and instant, reducing both cost and latency.
• Per-cell provenance — every value traces back to its source with confidence scores and reasoning.
• Batch mode — process large backlogs at 50% cost with a 48-hour delivery window.
• 12 source connectors: Google Drive, Gmail, SharePoint, OneDrive, Outlook, Teams, Notion, HubSpot, Granola AI, SQL databases, Amazon S3, and Azure Blob Storage.
• 7 delivery connectors: webhook (HMAC-SHA256), SFTP, Amazon S3, Azure Blob, Google Drive, OneDrive, and Google Sheets for pushing results downstream.

Quick Jobs use a cascading design of four Job phases that minimizes both cost and latency. Phase 1 fills cells using deterministic graph lookups at zero AI cost. Phase 2 deploys the Planner only for cells that Phase 1 could not resolve. Phase 3 re-runs lookups on AI output to normalize values to canonical codes. Phase 4 applies transforms, validates constraints, and fills remaining gaps with full grid context. This approach ensures that the cheapest, fastest method always runs first.

Quick API Example

Talonic exposes a full REST API so you can integrate document extraction into any workflow programmatically. The simplest way to get started is to upload a document via the /v1/sources/:sourceId/documents endpoint. The platform automatically runs OCR, classification, and extraction — you poll for results or receive a webhook notification when processing completes.

curl -X POST https://api.talonic.com/v1/sources/src_abc123/documents \
  -H "Authorization: Bearer $TALONIC_API_KEY" \
  -F "file=@invoice.pdf" \
  -F "processing_mode=realtime"

{
  "document_id": "doc_7f3a1b2c",
  "filename": "invoice.pdf",
  "status": "queued",
  "processing_mode": "realtime",
  "source_id": "src_abc123"
}

Once a document is processed, you can retrieve its captured fields using the document fields endpoint. Each field includes its canonical name from the Field Registry, the extracted value, a confidence score between 0 and 1, and the field's maturity level (the numeric tier in the response). This makes every extraction result fully auditable and traceable back to the knowledge graph concept it belongs to.

curl https://api.talonic.com/v1/documents/doc_7f3a1b2c/fields \
  -H "Authorization: Bearer $TALONIC_API_KEY"