/api/v1/context

Structured Corporate
Context for RAG & Agents.

Don't waste your context window on raw filings, article noise, or messy HTML. Retrieve tagged annual accounts sections, company news signals, and website context sections through a deterministic API.

Read API Docs View Sample Payload

Raw Documents are Toxic to AI Agents.

Feeding an autonomous agent a raw 80-page Annual Report or a messy HTML scrape is inefficient. It burns tokens, increases latency, and confuses the model with irrelevant boilerplate.

To build reliable financial agents, you don't need documents; you need structured narrative segments. We have already done the heavy lifting of extracting, cleaning, and categorizing the messy reality of UK corporate data so your system doesn't have to.

The Context Index

/api/v1/context/accounts · /api/v1/context/news · /api/v1/context/website

Statutory Filings

● Risk Profiles: Isolated "Principal Risks and Uncertainties" sections.
● Narrative: Directors' Reports and Strategic Reports.
● Audit Context: Independent auditor notes.

Web Footprint

Live

● Identity: "About Us", "History", "Our Story" blocks.
● Commercial: "Products", "Services", "Pricing" blocks.
● Operational: "Contact", "Careers", "Team" blocks.

News & Signals

● Expansion: Geographic expansion and market entries.
● M&A: Acquisitions, mergers, divestitures.
● Capital: Major investment, insolvency, and restructuring events.

From Unstructured Sources to JSON-Ready.

Step 01

Ingestion

Monitoring UK entities across annual accounts, company-level news, and tagged website sections.

Step 02

OCR & Vision

Proprietary vision models reconstruct layout & hierarchy.

Step 03

Classification

Semantic tagging by section and event category (e.g. principal_risks_and_uncertainties).

Step 04

Delivery

You query specific filing sections, website sections, or articles via API, not whole source files.

Precision Retrieval.

Fetch only the sections you need. Retrieve targeted accounts sections such as "Principal Risks and Uncertainties" or "Directors' Report" instead of parsing entire annual reports.

Live Index

OCR Verified

POST /api/v1/context/accounts

$ curl -X POST https://data.zint.io/api/v1/context/accounts \
  -H "Authorization: Bearer ZINT_KEY" \
  -H "Content-Type: application/json" \
  -d '{"company_number":"09876543","sections":["principal_risks_and_uncertainties","directors_report"],"num_filings":1}'


{
  "credits_charged": 10,
  "preview_only": false,
  "result": {
    "company_number": "09876543",
    "found": true,
    "filings": [
      {
        "made_up_to": "2024-12-31",
        "metadata": {
          "description": 2,
          "num_pages": 42,
          "parsing_failed": false
        },
        "available_sections": [
          "Strategic Report",
          "Principal Risks and Uncertainties",
          "Directors' Report",
          "Principal Activities"
        ],
        "sections": [
          {
            "category": "Principal Risks and Uncertainties",
            "parent_category": "Strategic Report",
            "text": "The company faces significant exposure to input cost volatility due to...",
            "start_page": 7,
            "end_page": 9
          },
          {
            "category": "Principal Activities",
            "parent_category": "Directors' Report",
            "text": "The principal activity of the company during the year was...",
            "start_page": 11,
            "end_page": 12
          }
        ]
      }
    ]
  }
}

Don't want to build the reasoning layer?

If you prefer the answer rather than the source material, use our Magic Insights API.

Explore Magic Insights →

Ground your Agent in Reality.

Start building with the UK's deepest contextual index today.

Get API Key

Structured Corporate Context for RAG & Agents.