Skip to main content

​
Overview

Projects are the top-level organizational unit in Mimir AIP. Each project groups all related resources for a specific use case or domain, providing isolation and management for:
  • Pipelines: Data ingestion and processing workflows
  • Ontologies: Domain vocabulary and entity definitions
  • ML Models: Machine learning models for predictions
  • Digital Twins: Real-time entity graphs and simulations
  • Storage Configs: Backend connections for data persistence
  • Schedules: Recurring pipeline execution triggers
Projects are fully isolated. Resources in one project cannot directly reference resources in another project.

​
Project Structure

pkg/models/project.go:15
type Project struct {
    ID          string
    Name        string
    Description string
    Version     string            // Semantic version (e.g., "1.0.0")
    Status      ProjectStatus     // active, archived, draft
    Metadata    ProjectMetadata
    Components  ProjectComponents // Resource associations
    Settings    ProjectSettings
}

type ProjectComponents struct {
    Pipelines      []string
    Ontologies     []string
    MLModels       []string
    DigitalTwins   []string
    StorageConfigs []string
}

​
Creating a Project

curl -X POST http://localhost:8080/api/projects \
  -H "Content-Type: application/json" \
  -d '{
    "name": "customer-analytics",
    "description": "Customer behavior analysis platform",
    "version": "1.0.0",
    "status": "active",
    "tags": ["analytics", "customer"],
    "settings": {
      "timezone": "America/New_York",
      "environment": "production"
    }
  }'

​
Validation Rules

  • Pattern: ^[a-zA-Z0-9_-]{3,50}$
  • Must be 3-50 characters
  • Alphanumeric, hyphens, and underscores only
  • Must be unique across non-archived projects
pkg/project/service.go:15
var projectNameRegex = regexp.MustCompile(`^[a-zA-Z0-9_-]{3,50}$`)
  • Pattern: ^\d+\.\d+\.\d+$
  • Must follow semantic versioning (e.g., “1.0.0”)
  • Defaults to “1.0.0” if not provided
pkg/project/service.go:16
var versionRegex = regexp.MustCompile(`^\d+\.\d+\.\d+$`)
Valid status values:
  • active — Project is operational (default)
  • draft — Project is being configured
  • archived — Read-only, hidden from listings
Archived projects maintain their data but cannot be modified.

​
Listing Projects

curl http://localhost:8080/api/projects
Query Parameters:
ParameterTypeDescription
statusstringFilter by project status
limitintegerNumber of results to return
offsetintegerNumber of results to skip

​
Updating a Project

curl -X PATCH http://localhost:8080/api/projects/proj-uuid-1234 \
  -H "Content-Type: application/json" \
  -d '{
    "description": "Updated customer analytics platform",
    "version": "1.1.0",
    "tags": ["analytics", "customer", "ml"]
  }'
Only provided fields are updated. Omitted fields remain unchanged.

​
Cloning a Project

Create a copy of an existing project with a new name. The clone includes all settings and tags, but components are empty. 
curl -X POST http://localhost:8080/api/projects/proj-uuid-1234/clone \
  -H "Content-Type: application/json" \
  -d '{"new_name": "customer-analytics-dev"}'
Cloned project characteristics:
  • New unique ID assigned
  • Description appended with ” (cloned)”
  • Version reset to “1.0.0”
  • Status set to draft
  • All component arrays empty
  • Settings and tags copied from original
pkg/project/service.go:214
cloned := &models.Project{
    ID:          uuid.New().String(),
    Name:        newName,
    Description: original.Description + " (cloned)",
    Version:     "1.0.0",
    Status:      models.ProjectStatusDraft,
    // Components arrays are empty
    Settings:    original.Settings,
}

​
Managing Components

Associate resources with a project using the component management endpoints. 
curl -X POST http://localhost:8080/api/projects/proj-uuid-1234/pipelines/pipe-uuid-5678
Pipelines are automatically associated when created with project_id, but can also be manually linked.
curl -X DELETE http://localhost:8080/api/projects/proj-uuid-1234/pipelines/pipe-uuid-5678
Removes the association. The pipeline itself is not deleted.
The same pattern applies for:
  • Ontologies: /ontologies/:id
  • ML Models: /ml-models/:id
  • Digital Twins: /digital-twins/:id
  • Storage Configs: /storage/:id
pkg/project/service.go:242
func (s *Service) AddPipeline(projectID, pipelineID string) error {
    project, err := s.store.GetProject(projectID)
    if err != nil {
        return err
    }
    
    if slices.Contains(project.Components.Pipelines, pipelineID) {
        return nil // Already associated
    }
    
    project.Components.Pipelines = append(project.Components.Pipelines, pipelineID)
    project.Metadata.UpdatedAt = time.Now()
    
    return s.store.SaveProject(project)
}

​
Deleting a Project

Deleting a project performs a soft delete by marking it as archived. 
curl -X DELETE http://localhost:8080/api/projects/proj-uuid-1234
Deleting a project does not delete its associated resources (pipelines, models, etc.). Those must be deleted separately.
pkg/project/service.go:177
func (s *Service) Delete(id string) error {
    project, err := s.store.GetProject(id)
    if err != nil {
        return err
    }
    
    // Soft delete - update status to archived
    project.Status = models.ProjectStatusArchived
    project.Metadata.UpdatedAt = time.Now()
    
    return s.store.SaveProject(project)
}

​
Project Settings

Timezone

Controls timestamp display and schedule execution. Defaults to "UTC".
{"timezone": "America/New_York"}

Environment

Deployment context label. Defaults to "development".
{"environment": "production"}

​
Best Practices

Use descriptive, kebab-case names:
  • customer-analytics
  • supply-chain-optimization
  • iot-sensor-monitoring
Avoid generic names like project1 or test.
Increment versions when making significant changes:
  • Patch (1.0.1): Bug fixes, minor updates
  • Minor (1.1.0): New features, backwards compatible
  • Major (2.0.0): Breaking changes
Use tags for categorization and filtering:
  • Domain: analytics, iot, ml
  • Team: data-team, ml-team
  • Purpose: production, experiment
Recommended workflow:
  1. Create as draft while configuring
  2. Activate when ready: active
  3. Archive when no longer needed: archived
Never delete without archiving first.

​
Example: Complete Project Setup

# 1. Create project
curl -X POST http://localhost:8080/api/projects \
  -H "Content-Type: application/json" \
  -d '{
    "name": "customer-analytics",
    "description": "Customer behavior analysis",
    "version": "1.0.0",
    "status": "draft",
    "tags": ["analytics", "customer"]
  }'

# 2. Create storage config
curl -X POST http://localhost:8080/api/storage/configs \
  -H "Content-Type: application/json" \
  -d '{
    "project_id": "proj-uuid-1234",
    "plugin_type": "postgres",
    "config": {"connection_string": "postgres://..."}
  }'

# 3. Create ontology
curl -X POST http://localhost:8080/api/ontologies \
  -H "Content-Type: application/json" \
  -d '{
    "project_id": "proj-uuid-1234",
    "name": "customer-ontology",
    "content": "@prefix : <http://example.org/customer#> ..."
  }'

# 4. Create pipeline
curl -X POST http://localhost:8080/api/pipelines \
  -H "Content-Type: application/json" \
  -d '{
    "project_id": "proj-uuid-1234",
    "name": "customer-import",
    "type": "ingestion",
    "steps": [...]
  }'

# 5. Activate project
curl -X PATCH http://localhost:8080/api/projects/proj-uuid-1234 \
  -H "Content-Type: application/json" \
  -d '{"status": "active"}'

​
Next Steps

Pipelines

Create data ingestion and processing workflows.

Ontologies

Define your domain vocabulary and entity types.

Storage

Configure storage backends for your data.

Build docs developers (and LLMs) love

Get started for freeTalk to us