Feat: organization models

[hiento09]

This pull request introduces a comprehensive set of changes to support AI model management and Kubernetes integration in the API gateway. The main additions are new domain models for AI models, a Kubernetes service for cluster and GPU management, and the necessary dependency injection wiring to expose these features via HTTP APIs. The changes are grouped below by theme.

AI Model Management:

• Added a new models domain package (apps/jan-api-gateway/application/app/domain/organization/models/model.go) defining types for model creation, resource requirements, model status, filtering, and cluster/GPU resource summaries. This enables structured management of AI models within organizations.

Kubernetes Integration:

• Introduced a new KubernetesService (apps/jan-api-gateway/application/app/infrastructure/kubernetes/kubernetes_service.go) that provides cluster connectivity, CRD checks, GPU node/resource discovery, and storage class validation. This service is essential for deploying and managing models on Kubernetes clusters.

Dependency Injection Wiring:

• Registered the new ModelService and Kubernetes-related services (NewKubernetesService, NewModelDeploymentManager) in the service and infrastructure provider sets, making them available for use throughout the application. [1] [2]

HTTP API Exposure:

• Added new HTTP route providers for models and Kubernetes APIs, enabling external access to model management and cluster status endpoints.

General Integration:

• Updated imports in the relevant provider and route files to include the new model and Kubernetes modules, ensuring all new functionality is properly wired into the application. [1] [2]

Issue: #128

[hiento09]

Example: request

curl -X 'POST' \
  'http://localhost:64185/v1/organization/models' \
  -H 'accept: application/json' \
  -H 'Authorization: Bearer *****' \
  -H 'Content-Type: application/json' \
  -d '{
  "command": [
    "sh", "-c", "python3 -m vllm.entrypoints.openai.api_server --host 0.0.0.0 --port 8000 --uvicorn-log-level warning --model janhq/Jan-v1-2509 --served-model-name jan-v1-4b --max-num-batched-tokens 16384 --enable-auto-tool-choice --max-model-len 131072 --tool-call-parser hermes --reasoning-parser qwen3 --compilation-config '\''{\"cudagraph_mode\":\"FULL_AND_PIECEWISE\",\"compile_sizes\":[1,2,4]}'\'' --async-scheduling --api-server-count 4"
  ],
  "description": "jan-v1-4b",
  "display_name": "jan-v1-4b",
  "gpu_count": 1,
  "image": "registry.menlo.ai/dockerhub/vllm/vllm-openai:v0.10.2",
  "initial_delay_seconds": 600,
  "name": "jan-v1-4b",
  "replicas": 1,
  "storage_size": 30,
  "tags": []
}'

Response:

{
  "model": {
    "id": "jan-v1-4b",
    "organization_id": 1,
    "display_name": "jan-v1-4b",
    "description": "jan-v1-4b",
    "status": "creating",
    "version": "",
    "requirements": {
      "cpu": "1",
      "memory": "2Gi",
      "gpu": {
        "min_vram": "8Gi",
        "preferred_vram": "16Gi",
        "gpu_type": "nvidia",
        "min_gpus": 1,
        "max_gpus": 1
      }
    },
    "namespace": "jan-models",
    "deployment_name": "jan-v1-4b",
    "service_name": "jan-v1-4b",
    "tags": [],
    "managed": true,
    "created_at": "2025-10-02T04:24:27.173970526Z",
    "updated_at": "2025-10-02T04:24:27.173970831Z",
    "created_by_user_id": "user_igjyupzhnmh56bh9x3hj2v80"
  }
}

[hiento09]

Close due to domain route conflict