Quickstart - ARGOS Identity

You will build a workflow that verifies three documents a vendor submitted, then create a profile and get an analysis result through the API.

How to Use

You tell your Review Ops Agent how the work is done in the dashboard, and run it for each subject through the API.

Omni process flow diagram — Configuration (policy, engines, output schema) → data ingestion (OCR, normalization) → analysis and validation → report, JSON, and audit log generation → delivery to your server.

Building blocks

Plain-language policy

Write how the work is done and your decision criteria as ordinary sentences. One policy clause becomes one execution step.

Automatic engine assignment

You never pick which engine a check needs. Engines are attached from the content of the policy, and the execution order is decided for you.

Output schema

Define the JSON structure you receive. Results come back in exactly the shape your system expects, so no post-processing conversion is needed.

Grounds for every verdict

You get more than values — you get why each value came out that way, with a step-by-step execution record and every external lookup in the audit log.

The whole flow

Configure the workflow (dashboard)

Write the policy and define the output schema you want back. Saving generates the playbook.

Create a profile (API)

Create one profile per verification subject. One vendor or one applicant is one profile.

Add documents (API)

Upload documents to the profile. Images and PDFs go through OCR; office documents go through text extraction.

Run the analysis (API)

The playbook runs in order, and you get back the verdict along with results shaped by your output schema.

Creating and editing workflows happens only in the dashboard. Everything else can be automated through the API. You run step 1 yourself in section 1 below, and steps 2–4 in section 2.

1. Create a project and a workflow in the dashboard

Create a project

In the dashboard, select Create Project and enter a name and description. This documentation uses a project named Demo as its example.An API key is issued along with the project. Find it on the API Key tab in the project settings — managing your API key.

Create a workflow

Inside the project, select Create New Workflow to open the four-step wizard.

Basic information — name it Vendor Onboarding Document Verification Policy
Policy definition — write, in plain language, what you verify and against which criteria
AI model — without a choice, the default model is used
Output schema — define the JSON structure you want back from an analysis

Saving generates a playbook (the execution plan) from the policy text. It takes 2–3 minutes. The full procedure is in creating a workflow.

You now have the two values the rest of this page needs — an API key and a workflow ID (wf_ prefix). You can also find the workflow ID by reading the project.

curl "https://client-omni-api.argosidentity.com/v1/projects/proj_75hgxoxamzn9" \
  -H "x-api-key: your-api-key-here"

2. Run an analysis through the API

Create a profile

A profile is the unit that holds one verification subject. One vendor or one applicant is one profile.

curl -X POST "https://client-omni-api.argosidentity.com/v1/workflows/wf_3p8k2m5r9v1t/profiles" \
  -H "x-api-key: your-api-key-here" \
  -H "Content-Type: application/json" \
  -d '{
    "name": "example",
    "description": "Example profile for document verification",
    "metadata": { "vendor_id": "VND-2026-0142" }
  }'

id in the response is the profile ID, and folders[0].id is the default folder created with it. You need both in the next step.

{
  "id": "pf_6h4y8s2f0v9n",
  "folders": [{ "id": "fd_9pq2mklt83bx", "name": "default", "isDefault": true }]
}

Add the documents

Upload the documents to the default folder. You can send a file or plain text.

curl -X POST "https://client-omni-api.argosidentity.com/v1/folders/fd_9pq2mklt83bx/items/upload" \
  -H "x-api-key: your-api-key-here" \
  -F "file=@/path/to/certificate-of-good-standing.pdf" \
  -F "name=certificate of good standing" \
  -F "metadata={\"doc_type\":\"certificate_of_good_standing\"}"

Record the id (item_ prefix) from the response — you pass it in the analysis request. Add all three documents the same way.

A single file is limited to 10MB, and one profile to 100MB of uploads in total. There is no per-folder count limit. The supported formats are in adding an item, and a format missing from that list is rejected at upload.

Wait for processing to finish

An added item goes through text extraction and moves from PENDING to ACTIVE. Every item must be ACTIVE before an analysis can run.

curl "https://client-omni-api.argosidentity.com/v1/profiles/pf_6h4y8s2f0v9n/items/status" \
  -H "x-api-key: your-api-key-here"

{
  "profileId": "pf_6h4y8s2f0v9n",
  "isRagReady": true,
  "summary": { "total": 3, "pending": 0, "active": 3, "failed": 0 },
  "items": [
    {
      "id": "item_axfkh2noof85",
      "name": "certificate of good standing",
      "type": "file",
      "status": "ACTIVE",
      "processedAt": "2026-09-10T05:35:38.902Z"
    }
  ]
}

Poll once per second until isRagReady is true, and stop at 60 seconds. An item that is still not ACTIVE by then is faster to delete and upload again.

Run the analysis

Pass the item IDs to include as targetItems. The workflow’s policy, engines, and output schema are applied automatically.

curl -X POST "https://client-omni-api.argosidentity.com/v1/analyses/pf_6h4y8s2f0v9n/analyze" \
  -H "x-api-key: your-api-key-here" \
  -H "Content-Type: application/json" \
  -d '{
    "targetItems": [
      { "itemId": "item_axfkh2noof85" },
      { "itemId": "item_lxzzl2vedy9y" },
      { "itemId": "item_wfepr5xxo6w2" }
    ]
  }'

Analyses are asynchronous. The response returns only the analysis ID, immediately.

{ "id": "analysis_lsc7icgm9cjd" }

Read the result

Poll the status with the analysis ID, and read the result once status is completed.

curl "https://client-omni-api.argosidentity.com/v1/analyses/analysis_lsc7icgm9cjd" \
  -H "x-api-key: your-api-key-here"

{
  "status": "completed",
  "verificationStatus": "verified",
  "confidence": { "score": 98, "level": "HIGH" },
  "extractedData": {
    "company": {
      "legal_name": {
        "value": "ACME TRADING LLC",
        "reasoning": "Read from the entity name field on the certificate of good standing.",
        "sourceStep": 1
      }
    }
  }
}

There are three things to read.

Field	What it is
`verificationStatus`	The verdict for the whole analysis — `verified` · `pending_review` · `rejected`
`confidence.score`	The Analysis Score. How completely the output was filled in (0–100)
`extractedData`	The values extracted per your output schema. Leaf fields are wrapped in a `value` · `reasoning` · `sourceStep` envelope

When you need a PDF for a person to read, get it from GET /analyses/:analysisId/report.

Next steps

Reading analysis results

The envelope structure, why a field is empty, and the order to read the verdicts in.

Core concepts

The data hierarchy from project to analysis, and what a playbook is.

Creating a workflow

The whole dashboard procedure, from writing the policy to the output schema.

API reference

Full definitions for all 23 endpoints.

​How to Use

​Building blocks

Plain-language policy

Automatic engine assignment

Output schema

Grounds for every verdict

​The whole flow

​1. Create a project and a workflow in the dashboard

​2. Run an analysis through the API

​Next steps

Reading analysis results

Core concepts

Creating a workflow

API reference

How to Use

Building blocks

The whole flow

1. Create a project and a workflow in the dashboard

2. Run an analysis through the API

Next steps