Projects

Group models into projects and read the roll-ups across them: traffic, alerts and the inputs each model relies on.

A project is a folder of models with roll-ups across them: traffic, alerts, the frameworks and tasks in use, and which inputs drive each model. Models join a project through project_id on create or update. Deleting a project leaves its models in the workspace, ungrouped.

#The Project object

summary is null unless you ask for it with include=summary.

Attributes

idstring

prj_ identifier.

name / slug / descriptionstring

model_countinteger

Models in the project.

summaryobject | null

With include=summary: active_models, lineages (distinct models, versions counted once), frameworks and tasks counts, predictions_total, explanations_total, predictions_30d, explanations_30d, alerts_high, alerts_medium (from each model's latest insight per type), analyses_running, last_activity_at and a daily series of the last 31 days.

created_at / updated_attimestamp

The Project object, with include=summary

{
  "object": "project",
  "id": "prj_40c081b2465e96db2cdc",
  "name": "Risk platform",
  "slug": "risk-platform",
  "description": "Models behind the lending decision.",
  "model_count": 3,
  "created_at": "2026-10-31T22:40:22.963721Z",
  "updated_at": "2026-10-31T22:40:22.963721Z",
  "summary": {
    "active_models": 3,
    "lineages": 2,
    "frameworks": {"sklearn": 2, "xgboost": 1},
    "tasks": {"binary_classification": 3},
    "predictions_total": 1560,
    "explanations_total": 27,
    "predictions_30d": 1136,
    "explanations_30d": 19,
    "alerts_high": 0,
    "alerts_medium": 1,
    "analyses_running": 0,
    "last_activity_at": "2026-10-31T22:40:25Z",
    "series": [
      {
        "date": "2026-10-01",
        "predictions": 0,
        "explanations": 0
      },
      …,
      {
        "date": "2026-10-31",
        "predictions": 212,
        "explanations": 6
      }
    ]
  }
}

#List projects

GET/v1/projectsAPI key or session

Projects in your organization, by name.

Query parameters

qstringoptional

Match on name or slug.

includestringoptional

summary to add each project's roll-up. Unknown values are rejected (400 invalid_include).

limitintegeroptionaldefault 50

1 to 200.

offsetintegeroptionaldefault 0

Returns

A page of Projects.

for project in client.projects.list(include="summary").data:
    s = project.summary
    print(project.slug, project.model_count, s["predictions_30d"], s["alerts_high"], s["alerts_medium"])

curl "https://api.kaimei.ai/v1/projects?include=summary" \
  -H "Authorization: Bearer $KAIMEI_API_KEY"

200 OKResponse

{
  "object": "list",
  "data": [
    {
      "object": "project",
      "id": "prj_40c081b2465e96db2cdc",
      "name": "Risk platform",
      "slug": "risk-platform",
      "description": "Models behind the lending decision.",
      "model_count": 3,
      "created_at": "2026-10-31T22:40:22.963721Z",
      "updated_at": "2026-10-31T22:40:22.963721Z",
      "summary": {
        "active_models": 3,
        "lineages": 2,
        "frameworks": {"sklearn": 2, "xgboost": 1},
        "tasks": {"binary_classification": 3},
        "predictions_total": 1560,
        "explanations_total": 27,
        "predictions_30d": 1136,
        "explanations_30d": 19,
        "alerts_high": 0,
        "alerts_medium": 1,
        "analyses_running": 0,
        "last_activity_at": "2026-10-31T22:40:25Z",
        "series": …
      }
    }
  ],
  "total": 1,
  "limit": 50,
  "offset": 0,
  "has_more": false
}

#Create a project

POST/v1/projectsAPI key or session

Create an empty project. Add models with project_id on POST /v1/models or PATCH /v1/models/{model}.

Body parameters

namestringrequired

1 to 120 characters.

slugstringoptional

Unique within the organization (409 slug_taken). Derived from the name when omitted.

descriptionstringoptional

Returns

The Project. 201.

project = client.projects.create(name="Risk platform", description="Models behind the lending decision.")
client.models.assign_project("credit-risk-v3", project.id)

curl -X POST https://api.kaimei.ai/v1/projects \
  -H "Authorization: Bearer $KAIMEI_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{"name": "Risk platform", "description": "Models behind the lending decision."}'

201 CreatedResponse

{
  "object": "project",
  "id": "prj_40c081b2465e96db2cdc",
  "name": "Risk platform",
  "slug": "risk-platform",
  "description": "Models behind the lending decision.",
  "model_count": 0,
  "created_at": "2026-10-31T22:40:22.963721Z",
  "updated_at": "2026-10-31T22:40:22.963721Z",
  "summary": null
}

#Retrieve a project

GET/v1/projects/{project}API key or session

One project by id or slug.

Path parameters

projectstringrequired

Project id or slug.

Query parameters

includestringoptional

summary to add the roll-up.

Returns

A Project.

project = client.projects.retrieve("risk-platform", include="summary")

curl https://api.kaimei.ai/v1/projects/risk-platform?include=summary \
  -H "Authorization: Bearer $KAIMEI_API_KEY"

200 OKResponse

{
  "object": "project",
  "id": "prj_40c081b2465e96db2cdc",
  "name": "Risk platform",
  "slug": "risk-platform",
  "description": "Models behind the lending decision.",
  "model_count": 3,
  "created_at": "2026-10-31T22:40:22.963721Z",
  "updated_at": "2026-10-31T22:40:22.963721Z",
  "summary": {
    "active_models": 3,
    "lineages": 2,
    "frameworks": {"sklearn": 2, "xgboost": 1},
    "tasks": {"binary_classification": 3},
    "predictions_total": 1560,
    "explanations_total": 27,
    "predictions_30d": 1136,
    "explanations_30d": 19,
    "alerts_high": 0,
    "alerts_medium": 1,
    "analyses_running": 0,
    "last_activity_at": "2026-10-31T22:40:25Z",
    "series": …
  }
}

#Project dashboard

GET/v1/projects/{project}/overviewAPI key or session

Everything the project page shows in one call: the project with its summary, every model in it, per-model 30-day traffic with the latest drift status, open medium and high alerts, the feature matrix (which inputs drive each model over its recent explanations, with the features shared by two or more models), and the most recent analyses and explanations.

Path parameters

projectstringrequired

Project id or slug.

Returns

A project.overview: project, models, model_stats[] (model_id, predictions_30d, explanations_30d, avg_latency_ms, alerts_high, alerts_medium, drift_status, drift_psi), alerts[] (Insights), feature_matrix {features, rows[] {model_id, explanations, shares}, shared}, recent_analyses[] and recent_explanations[].

overview = client.projects.overview("risk-platform")
for stat in overview.model_stats:
    print(stat["model_id"], stat["predictions_30d"], stat["drift_status"], stat["drift_psi"])
matrix = overview.feature_matrix
print(matrix["shared"])                       # inputs two or more models use
for row in matrix["rows"]:
    print(row["model_id"], {k: v for k, v in row["shares"].items() if v})

curl https://api.kaimei.ai/v1/projects/risk-platform/overview \
  -H "Authorization: Bearer $KAIMEI_API_KEY"

200 OKResponse

{
  "object": "project.overview",
  "project": {
    "object": "project",
    "id": "prj_40c081b2465e96db2cdc",
    "name": "Risk platform",
    "slug": "risk-platform",
    "description": "Models behind the lending decision.",
    "model_count": 3,
    "created_at": "2026-10-31T22:40:22.963721Z",
    "updated_at": "2026-10-31T22:40:22.963721Z",
    "summary": {
      "active_models": 3,
      "lineages": 2,
      "frameworks": {"sklearn": 2, "xgboost": 1},
      "tasks": {"binary_classification": 3},
      "predictions_total": 1560,
      "explanations_total": 27,
      "predictions_30d": 1136,
      "explanations_30d": 19,
      "alerts_high": 0,
      "alerts_medium": 1,
      "analyses_running": 0,
      "last_activity_at": "2026-10-31T22:40:25Z",
      "series": …
    }
  },
  "models": [
    {
      "object": "model",
      "id": "mdl_390bd6d9568f626851ef",
      "slug": "credit-risk-v3",
      …: …
    },
    …
  ],
  "model_stats": [
    {
      "model_id": "mdl_390bd6d9568f626851ef",
      "predictions_30d": 860,
      "explanations_30d": 19,
      "avg_latency_ms": 31.4,
      "alerts_high": 0,
      "alerts_medium": 1,
      "drift_status": "moderate",
      "drift_psi": 0.148039
    },
    …
  ],
  "alerts": [
    {
      "object": "insight",
      "id": "ins_a0a2b5c4a5ddb552758a",
      "type": "drift",
      "severity": "medium",
      "summary": "Recent traffic has moderate drift (PSI 0.15). Shifted inputs: `loan_amount`, `debt_to_income`, `income`.",
      …: …
    }
  ],
  "feature_matrix": {
    "features": [
      "credit_score", "debt_to_income", "income", "utilization",
      "num_delinquencies", …
    ],
    "rows": [
      {
        "model_id": "mdl_390bd6d9568f626851ef",
        "explanations": 19,
        "shares": {
          "credit_score": 0.2691,
          "income": 0.1586,
          "utilization": 0.1402,
          "debt_to_income": 0.1218,
          "num_delinquencies": 0.0934,
          …: …
        }
      },
      {
        "model_id": "mdl_c48c5582152e44e34e64",
        "explanations": 8,
        "shares": {
          "credit_score": 0.4681,
          "utilization": 0.2979,
          "income": 0.1277,
          "num_delinquencies": 0.1064,
          "debt_to_income": null
        }
      }
    ],
    "shared": [
      "credit_score", "income", "utilization",
      "num_delinquencies"
    ]
  },
  "recent_analyses": [
    {
      "object": "analysis",
      "id": "anl_b8d673126dcf3cdbe79a",
      "type": "full",
      "status": "completed",
      …: …
    }
  ],
  "recent_explanations": [
    {
      "object": "explanation",
      "id": "exp_7d7add16c796eafac087",
      "method": "shap",
      …: …
    },
    …
  ]
}

#Update a project

PATCH/v1/projects/{project}API key or session

Rename, re-slug or describe a project.

Path parameters

projectstringrequired

Project id or slug.

Body parameters

namestringoptional

slugstringoptional

Must be unique (409 slug_taken).

descriptionstring | nulloptional

Returns

The Project.

client.projects.update("risk-platform", description="Credit decisioning models.")

curl -X PATCH https://api.kaimei.ai/v1/projects/risk-platform \
  -H "Authorization: Bearer $KAIMEI_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{"description": "Credit decisioning models."}'

200 OKResponse

{
  "object": "project",
  "id": "prj_40c081b2465e96db2cdc",
  "name": "Risk platform",
  "slug": "risk-platform",
  "description": "Credit decisioning models.",
  "model_count": 3,
  "created_at": "2026-10-31T22:40:22.963721Z",
  "updated_at": "2026-10-31T22:40:22.989821Z",
  "summary": null
}

#Delete a project

DELETE/v1/projects/{project}API key or session

Remove the folder. Models inside it stay in the workspace, ungrouped.

Path parameters

projectstringrequired

Project id or slug.

Returns

A deletion receipt.

client.projects.delete("risk-platform")

curl -X DELETE https://api.kaimei.ai/v1/projects/risk-platform \
  -H "Authorization: Bearer $KAIMEI_API_KEY"

200 OKResponse

{
  "id": "prj_40c081b2465e96db2cdc",
  "object": "project",
  "deleted": true
}