Document Analyzer

What You’ll Build

A workflow that:

Accepts a vault ID and search query via webhook
Searches your indexed documents
Analyzes the results with GPT-4o
Returns structured insights

Time to complete: 15 minutes

Architecture

┌──────────────┐     ┌──────────────┐     ┌──────────────┐
│   Webhook    │ ──▶ │ Vault Search │ ──▶ │   LLM        │
│   Trigger    │     │   (Hybrid)   │     │   Analysis   │
│              │     │              │     │              │
│  vaultId     │     │  10 chunks   │     │  Structured  │
│  query       │     │  + sources   │     │  response    │
└──────────────┘     └──────────────┘     └──────────────┘

Prerequisites

Case.dev API key (get one here)
A vault with ingested documents (we’ll set one up if you don’t have one)

Step 1: Set Up Your Vault

If you already have a vault with indexed documents, skip to Step 2.

Create a vault

cURL

export CASEDEV_API_KEY="sk_case_YOUR_API_KEY"

curl -X POST https://api.case.dev/vault \
  -H "Authorization: Bearer $CASEDEV_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{"name": "Legal Research Vault"}'

Save the returned id — you’ll need it.

Upload a document

cURL

# Get upload URL
UPLOAD_RESPONSE=$(curl -s -X POST https://api.case.dev/vault/YOUR_VAULT_ID/upload \
  -H "Authorization: Bearer $CASEDEV_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "filename": "contract.pdf",
    "contentType": "application/pdf"
  }')

# Extract the upload URL and object ID
UPLOAD_URL=$(echo $UPLOAD_RESPONSE | jq -r '.uploadUrl')
OBJECT_ID=$(echo $UPLOAD_RESPONSE | jq -r '.objectId')

# Upload your file
curl -X PUT "$UPLOAD_URL" \
  -H "Content-Type: application/pdf" \
  --data-binary @contract.pdf

echo "Object ID: $OBJECT_ID"

Ingest (index) the document

cURL

curl -X POST https://api.case.dev/vault/YOUR_VAULT_ID/ingest/$OBJECT_ID \
  -H "Authorization: Bearer $CASEDEV_API_KEY"

Ingestion is async. Wait for ingestionStatus: "completed" before searching. Poll the object endpoint to check status.

cURL

# Check status (poll until completed)
curl https://api.case.dev/vault/YOUR_VAULT_ID/objects/$OBJECT_ID \
  -H "Authorization: Bearer $CASEDEV_API_KEY" | jq '.ingestionStatus'

Step 2: Create the Workflow

Now let’s create the document analyzer workflow. Save this as document-analyzer.json:

JSON

{
  "name": "Document Analyzer",
  "description": "Search vault documents and generate AI analysis",
  "nodes": [
    {
      "id": "trigger-1",
      "type": "trigger",
      "label": "Webhook Trigger",
      "config": {
        "triggerType": "Webhook"
      }
    },
    {
      "id": "search-1",
      "type": "action",
      "label": "Vault Search",
      "config": {
        "actionType": "case-search/vault-search",
        "vaultId": "{{vaultId}}",
        "query": "{{query}}",
        "limit": 10,
        "searchMethod": "hybrid"
      }
    },
    {
      "id": "analyze-1",
      "type": "action",
      "label": "Analyze",
      "config": {
        "actionType": "case-llm/llm",
        "method": "chat",
        "model": "gpt-4o",
        "systemPrompt": "You are a legal document analyst. Analyze the provided document excerpts and answer the user's question. Always cite specific passages from the documents to support your analysis. If the documents don't contain relevant information, say so clearly.",
        "userPrompt": "## Document Excerpts\n\n{{results.Vault_Search.output.chunks}}\n\n## Sources\n{{results.Vault_Search.output.sources}}\n\n## Question\n{{query}}"
      }
    }
  ],
  "edges": [
    { "source": "trigger-1", "target": "search-1" },
    { "source": "search-1", "target": "analyze-1" }
  ]
}

Deploy the workflow

cURL

curl -X POST https://api.case.dev/workflows/v1/create \
  -H "Authorization: Bearer $CASEDEV_API_KEY" \
  -H "Content-Type: application/json" \
  -d @document-analyzer.json

Response:

Response

{
  "id": "wf_abc123xyz",
  "name": "Document Analyzer",
  "webhookUrl": "https://api.case.dev/workflows/v1/wf_abc123xyz/webhook",
  "webhookSecret": "whsec_aBcDeFgHiJkLmNoP",
  "status": "deployed",
  "message": "Workflow created and deployed successfully."
}

Save your webhook secret! It’s only shown once. You’ll need it to trigger executions.

Step 3: Execute the Workflow

Now trigger your workflow with a vault ID and query:

cURL

curl -X POST "https://api.case.dev/workflows/v1/wf_abc123xyz/webhook" \
  -H "X-Webhook-Secret: whsec_aBcDeFgHiJkLmNoP" \
  -H "Content-Type: application/json" \
  -d '{
    "vaultId": "YOUR_VAULT_ID",
    "query": "What are the key terms of this agreement?"
  }'

Async execution (default)

The default response returns an execution ID immediately:

Response

{
  "executionId": "wfex_execution123",
  "status": "running",
  "mode": "fire-and-forget"
}

Check status:

cURL

curl https://api.case.dev/workflows/v1/runs/wfex_execution123 \
  -H "Authorization: Bearer $CASEDEV_API_KEY"

Synchronous execution

For immediate results, add ?mode=sync:

cURL

curl -X POST "https://api.case.dev/workflows/v1/wf_abc123xyz/webhook?mode=sync" \
  -H "X-Webhook-Secret: whsec_aBcDeFgHiJkLmNoP" \
  -H "Content-Type: application/json" \
  -d '{
    "vaultId": "YOUR_VAULT_ID",
    "query": "What are the key terms of this agreement?"
  }'

Response includes the full analysis:

Response

{
  "status": "Completed",
  "duration": "8.42s",
  "output": {
    "results": {
      "Vault_Search": {
        "output": {
          "chunks": [...],
          "sources": [...]
        }
      },
      "Analyze": {
        "output": {
          "choices": [{
            "message": {
              "content": "Based on the contract excerpts, the key terms include:\n\n1. **Payment Terms**: Section 3.2 states that payment is due within 30 days...\n\n2. **Termination**: Either party may terminate with 90 days written notice (Section 7.1)...\n\n3. **Liability Cap**: Liability is limited to the total fees paid in the preceding 12 months (Section 9.3)..."
            }
          }]
        }
      }
    }
  }
}

Step 4: Access the Analysis

The LLM response is at:

output.results.Analyze.output.choices[0].message.content

Example extraction with jq:

cURL

curl -s https://api.case.dev/workflows/v1/runs/wfex_execution123 \
  -H "Authorization: Bearer $CASEDEV_API_KEY" \
  | jq -r '.output.results.Analyze.output.choices[0].message.content'

Extending the Workflow

Add PDF Report Generation

Generate a formatted PDF report from the analysis:

JSON

{
  "id": "format-1",
  "type": "action",
  "label": "Generate PDF",
  "config": {
    "actionType": "case-format/format",
    "content": "# Document Analysis Report\n\n**Query:** {{query}}\n\n## Analysis\n\n{{results.Analyze.output.choices[0].message.content}}\n\n## Sources\n\n{{results.Vault_Search.output.sources}}",
    "inputFormat": "md",
    "outputFormat": "pdf"
  }
}

Add edge: { "source": "analyze-1", "target": "format-1" }

Add Entity Extraction

Extract named entities before analysis:

JSON

{
  "id": "entities-1",
  "type": "action",
  "label": "Extract Entities",
  "config": {
    "actionType": "case-llm/llm",
    "method": "chat",
    "model": "gpt-4o",
    "systemPrompt": "Extract all named entities from the text. Return as JSON with arrays for: PERSON, ORGANIZATION, DATE, LOCATION, MONEY.",
    "userPrompt": "{{results.Vault_Search.output.chunks}}"
  }
}

Add Web Research

Supplement vault search with live web research:

JSON

{
  "id": "web-1",
  "type": "action",
  "label": "Web Research",
  "config": {
    "actionType": "case-search/web-search",
    "query": "{{query}} legal precedent case law",
    "numResults": 5
  }
}

Complete Enhanced Workflow

Here’s the full enhanced workflow with all extensions:

JSON

{
  "name": "Enhanced Document Analyzer",
  "description": "Search, extract entities, research, analyze, and generate PDF",
  "nodes": [
    {
      "id": "trigger",
      "type": "trigger",
      "label": "Webhook",
      "config": { "triggerType": "Webhook" }
    },
    {
      "id": "vault-search",
      "type": "action",
      "label": "Vault Search",
      "config": {
        "actionType": "case-search/vault-search",
        "vaultId": "{{vaultId}}",
        "query": "{{query}}",
        "limit": 15,
        "searchMethod": "hybrid"
      }
    },
    {
      "id": "web-search",
      "type": "action",
      "label": "Web Research",
      "config": {
        "actionType": "case-search/web-search",
        "query": "{{query}} legal analysis",
        "numResults": 5
      }
    },
    {
      "id": "extract-entities",
      "type": "action",
      "label": "Extract Entities",
      "config": {
        "actionType": "case-llm/llm",
        "method": "chat",
        "model": "gpt-4o",
        "systemPrompt": "Extract named entities as JSON: {people: [], organizations: [], dates: [], locations: [], monetary_amounts: []}",
        "userPrompt": "{{results.Vault_Search.output.chunks}}"
      }
    },
    {
      "id": "analyze",
      "type": "action",
      "label": "Comprehensive Analysis",
      "config": {
        "actionType": "case-llm/llm",
        "method": "chat",
        "model": "gpt-4o",
        "systemPrompt": "You are a senior legal analyst. Provide comprehensive analysis combining document excerpts with external research. Structure your response with: Executive Summary, Key Findings, Supporting Evidence, and Recommendations.",
        "userPrompt": "## Document Excerpts\n{{results.Vault_Search.output.chunks}}\n\n## Entities Found\n{{results.Extract_Entities.output.choices[0].message.content}}\n\n## External Research\n{{results.Web_Research.output.results}}\n\n## Question\n{{query}}"
      }
    },
    {
      "id": "generate-pdf",
      "type": "action",
      "label": "Generate Report",
      "config": {
        "actionType": "case-format/format",
        "content": "# Legal Analysis Report\n\n**Generated:** {{timestamp}}\n**Query:** {{query}}\n\n---\n\n{{results.Comprehensive_Analysis.output.choices[0].message.content}}\n\n---\n\n## Document Sources\n{{results.Vault_Search.output.sources}}",
        "inputFormat": "md",
        "outputFormat": "pdf"
      }
    }
  ],
  "edges": [
    { "source": "trigger", "target": "vault-search" },
    { "source": "trigger", "target": "web-search" },
    { "source": "vault-search", "target": "extract-entities" },
    { "source": "vault-search", "target": "analyze" },
    { "source": "web-search", "target": "analyze" },
    { "source": "extract-entities", "target": "analyze" },
    { "source": "analyze", "target": "generate-pdf" }
  ]
}

Production Tips

Error Handling

Check execution status before accessing results:

JavaScript

const run = await fetch(`/workflows/v1/runs/${executionId}`);
const data = await run.json();

if (data.status === 'Failed') {
  console.error('Workflow failed:', data.error);
  // Check data.steps for which step failed
}

Rate Limits

Vault search: 100 requests/minute
LLM calls: Based on your plan
Workflow executions: 1000/hour (default)

Cost Optimization

Component	Cost Driver	Optimization
Vault Search	Per query	Cache common queries
LLM Analysis	Token count	Use smaller `limit` values
PDF Generation	Per page	Summarize before formatting

Monitoring

View all executions in the Visual Builder:

Execution timeline
Step-by-step inputs/outputs
Error details with stack traces

Next Steps

Document Workflows Guide — More patterns and examples
Vault Search Reference — Advanced search options
API Reference — Full endpoint documentation

Cookbooks

​What You’ll Build

​Architecture

​Prerequisites

​Step 1: Set Up Your Vault

​Create a vault

​Upload a document

​Ingest (index) the document

​Step 2: Create the Workflow

​Deploy the workflow

​Step 3: Execute the Workflow

​Async execution (default)

​Synchronous execution

​Step 4: Access the Analysis

​Extending the Workflow

​Add PDF Report Generation

​Add Entity Extraction

​Add Web Research

​Complete Enhanced Workflow

​Production Tips

​Error Handling

​Rate Limits

​Cost Optimization

​Monitoring

​Next Steps

What You’ll Build

Architecture

Prerequisites

Step 1: Set Up Your Vault

Create a vault

Upload a document

Ingest (index) the document

Step 2: Create the Workflow

Deploy the workflow

Step 3: Execute the Workflow

Async execution (default)

Synchronous execution

Step 4: Access the Analysis

Extending the Workflow

Add PDF Report Generation

Add Entity Extraction

Add Web Research

Complete Enhanced Workflow

Production Tips

Error Handling

Rate Limits

Cost Optimization

Monitoring

Next Steps