1. Introduction

DocyAI is a drag-and-drop based intelligent document processing system. With a visual interface, users can quickly build, configure, and run document-handling Agents without writing code. These Agents automate tasks such as document parsing, data extraction, validation, and content generation.

Key Features:

• Drag-and-drop workflow builder, no coding required
• Supports multiple document formats (PDF, Word, images, etc.)
• Rich set of nodes: file input, data extraction, conditional logic, auditing, knowledge base integration, and more
• Multi-step and looped workflows supported

2. Getting Started

2.1 Sign Up & Log In

• Go to DocyAI
• Register with your email and password
• Log in to access the main dashboard

2.2 Create Your First Agent

1. Click “New Agent”
2. Enter the Agent name and description
3. An empty workflow canvas will be created

2.3 Build a Workflow (Example)

1. Drag Input File node → to upload the document you want to process
2. Drag Data Extractor node → to define fields to extract (e.g., Name, Date, Amount)
3. Drag Audit Node → to set validation rules (global or field-level)
4.  Mark the node result as the Task output.

👉 Tip: Connect nodes with lines to define execution order.

3. Teamspaces & Collaboration

3.1 What is a Teamspace?

A Teamspace in DocyAI is a shared workspace where you and your teammates can work together on Agents and document workflows.

3.2 Creating a Teamspace

1. From the dashboard, click “Create Teamspace”
2. Enter a Teamspace name
3. A new shared workspace will be created

3.3 Inviting Members

• Inside a Teamspace, go to “Members”
• Click “Invite” and enter the email address of the person you want to add
• The invited member can join and collaborate once they accept the invitation

3.4 Collaboration

• All members in the Teamspace can view and edit workflows
• Changes are saved in real time, making teamwork seamless

4. Features & Nodes

4.1 Node Types

• Input File: Upload or import files as workflow input.

  Every Agent must start with an Input node. It serves as the entry point for all documents and data.

• Data Extractor: Define extraction rules and capture structured data from documents

• Extract Signature: Detect and extract signature/seal information

• Audit Node – Global: Apply validation rules to the whole document

• Audit Node – Field: Validate specific fields

• Conditional Node: Branch the workflow based on conditions

• Interaction Node: Request user input or confirmation during the workflow.

  If the Interaction Node requires document input, it can create a loop, allowing new inputs to re-trigger the workflow.

• Knowledge Node: Query a knowledge base for answers or matches

• Document Template Node: Generate new documents based on templates

• Web Search Node: Fetch information from the internet

• Content Generator Node: Use AI to generate text content

4.2 Agent Operations

• Run Agent: Executes the workflow in sequence
• Debug Agent: Step-by-step run to troubleshoot issues
• Save & Publish:
  • Save as draft (private use only)
  • Publish to Marketplace (share with other users)

5. Common Use Cases

• Contract Review: Upload a contract → Extract fields (signing date, parties) → Apply audit rules (missing signature, date validity) → Output review report
• Invoice Processing: Upload invoice → Extract amount and tax ID → Validate invoice format → Export to Excel
• Knowledge Base Q\&A: Upload internal policy documents → Build Q\&A Agent → End users can ask “What’s the reimbursement process?” or “How does leave work?”

6. FAQ

Q1. Why did my node fail to run?

A: Input may be empty or fields not defined. Check upstream node configuration.

Q2. Why do I get errors uploading a PDF?

A: The file may exceed size limits or be in an unsupported format. Try converting to a compatible format.

Q3. How do I share my Agent?

A: Click “Publish” and select Marketplace to make it public.