# Projects

**Projects** are organizational containers that group related content based on business context or workflows. Unlike categories (which classify content by its properties), projects organize content by how it's being used—whether that's a research initiative, compliance audit, product launch, or client engagement.

## What is a Project?

A project is a named collection of content within a dataspace. Think of it as a folder or workspace that brings together documents relevant to a specific business objective.

**Key Characteristics**:

- **Business-focused**: Organize by initiative, not by document type
- **Many-to-many**: Content can belong to multiple projects simultaneously
- **Cross-category**: Projects can contain content from any category


**Example Project Structure**:


```
Project: "Loan Application - Smith, John (LA-2024-1547)"
├── Loan Application Form (Category: Application)
├── Credit Report (Category: Financial Document)
├── Employment Verification (Category: Verification Document)
├── Property Appraisal (Category: Appraisal Report)
└── Approval Memo (Category: Internal Memo)
```

All these documents have different categories, but they're all part of the same loan application process.

## Creating and Managing Projects

Projects support full CRUD operations (Create, Read, Update, Delete) via the Workspace API. You can create projects within a dataspace, list all projects, update project details (name and description), and delete projects when they're no longer needed.

**Note**: Deleting a project removes only the project container itself—the content it contains remains in the dataspace.

For complete API documentation, see the [Workspace API Reference](/api/workspace-openapi).

## Adding Content to Projects

Content can be associated with and dissociated from projects at any time. A single piece of content can belong to multiple projects.

### Associate Content with a Project


```bash
POST https://sws-{workspaceId}.cloud.syncdocs.ai/api/projects/{dataspaceId}/{projectId}/content/{contentId}
Authorization: Bearer <token>
```

**Example**: Add a credit report to the loan application project:


```bash
POST https://sws-{workspaceId}.cloud.syncdocs.ai/api/projects/sds-abc12345/proj-550e8400-uuid/content/cnt-12345678-uuid
Authorization: Bearer <token>
```

This creates an association between the project and the content. The content can still be associated with other projects.

### Remove Content from a Project


```bash
DELETE https://sws-{workspaceId}.cloud.syncdocs.ai/api/projects/{dataspaceId}/{projectId}/content/{contentId}
Authorization: Bearer <token>
```

This removes the association but does **not** delete the content itself.

### Get All Content in a Project


```bash
GET https://sws-{workspaceId}.cloud.syncdocs.ai/api/projects/{dataspaceId}/{projectId}/content
Authorization: Bearer <token>
```

**Response**:


```json
{
  "contentIds": [
    "cnt-12345678-uuid",
    "cnt-23456789-uuid",
    "cnt-34567890-uuid"
  ]
}
```

### Many-to-Many Relationships

Because projects use a many-to-many relationship with content, you can organize the same documents in multiple ways:

**Example**:


```
Content: "Credit Report - Smith, John"
├── Project: "Loan Application - Smith, John (LA-2024-1547)" (mortgage application)
├── Project: "Credit Review - October 2024" (monthly credit review batch)
└── Project: "Compliance Audit Q4 2024" (sample for audit review)
```

This flexibility allows different teams to organize the same content according to their workflows without duplication.

## Querying Project Content

Projects become especially powerful when combined with queries. You can scope queries to specific projects to focus AI responses on relevant context.

### Project-Scoped Query


```bash
POST https://sws-{workspaceId}.cloud.syncdocs.ai/api/content/{dataspaceId}/query
Authorization: Bearer <token>
Content-Type: application/json

{
  "query": "What is the applicant's employment status and annual income?",
  "context": {
    "contentFilters": {
      "projectId": "proj-550e8400-uuid"
    }
  }
}
```

**How This Works**:

1. Query filters content to only those in the specified project
2. Vector search operates on the filtered set
3. AI response is generated using only project-relevant documents
4. Citations reference documents within the project


### Combining Project and Other Filters

You can combine project filters with other content filters:


```bash
POST https://sws-{workspaceId}.cloud.syncdocs.ai/api/content/{dataspaceId}/query
Authorization: Bearer <token>
Content-Type: application/json

{
  "query": "What verification documents are still pending?",
  "context": {
    "contentFilters": {
      "projectId": "proj-550e8400-uuid",
      "categoryId": "cat-verification-uuid",
      "metadata": {
        "verificationStatus": "pending"
      }
    }
  }
}
```

This query searches only verification documents within the project that have pending status.

### Project-Wide Research Agent


```bash
POST https://sws-{workspaceId}.cloud.syncdocs.ai/api/content/{dataspaceId}/query
Authorization: Bearer <token>
Content-Type: application/json

{
  "query": "Provide a complete summary of this loan application and recommend approval or denial",
  "agentId": "agent-underwriting-uuid",
  "context": {
    "contentFilters": {
      "projectId": "proj-550e8400-uuid"
    }
  }
}
```

The underwriting agent analyzes all documents in the loan application project to provide a comprehensive recommendation.

## Next Steps

- [Understand Content](/concepts/content) - Learn how content relates to projects
- [Explore Queries](/concepts/queries) - Use projects to scope AI queries
- [Learn about Categories](/concepts/ontologies) - Compare with category-based organization
- [API Reference](/api) - Complete project API documentation