coreplexml 0.1.3

Python SDK for CorePlexML AutoML & MLOps platform

CorePlexML Python SDK

Official Python client for the CorePlexML AutoML & MLOps platform.

Installation

pip install coreplexml

Quick Start

from coreplexml import CorePlexMLClient

client = CorePlexMLClient(
    base_url="https://platform.coreplexml.io",
    api_key="your-api-key"
)

# List projects
projects = client.projects.list()

# Create a project
project = client.projects.create(name="My ML Project")

# Upload a dataset
dataset = client.datasets.upload(
    project_id=project["id"],
    file_path="data.csv",
    name="Training Data"
)

# Run an AutoML experiment
caps = client.experiments.capabilities()
print(caps["allowed_engines"], caps["execution_modes"])

experiment = client.experiments.create(
    project_id=project["id"],
    dataset_version_id=dataset["dataset_version_id"],
    target_column="label",
    problem_type="classification",
    engine="h2o",
    execution_mode="single",
    config={"max_models": 10, "max_runtime_secs": 300}
)

# Get the best model
models = client.models.list(project_id=project["id"])
best = models["items"][0]

# Deploy
deployment = client.deployments.create(
    project_id=project["id"],
    model_id=best["id"],
    name="production-v1"
)

# Predict
result = client.deployments.predict(
    deployment_id=deployment["id"],
    inputs={"feature1": 1.0, "feature2": "A"}
)
print(result["predictions"])

Resources

The client provides access to all platform resources:

Resource	Description
client.projects	Project CRUD
client.datasets	Dataset upload, versions, columns
client.experiments	AutoML experiment management
client.models	Model listing, metrics, explainability
client.deployments	Deploy models, predict, A/B tests
client.reports	Generate PDF/HTML reports
client.privacy	Privacy Suite (PII detection, compliance)
client.synthgen	Synthetic data generation (CTGAN, TVAE)
client.studio	What-If analysis sessions

Authentication

Create an API key from Profile > API Keys in the platform, then:

client = CorePlexMLClient(
    base_url="https://platform.coreplexml.io",
    api_key="cpx_your_api_key_here"
)

Error Handling

from coreplexml import CorePlexMLClient, NotFoundError, ValidationError

try:
    client.projects.get("nonexistent-id")
except NotFoundError:
    print("Project not found")
except ValidationError as e:
    print(f"Validation error: {e}")

Requirements

• Python >= 3.9
• requests >= 2.28

Plan-Aware AutoML

# Pro: choose one engine
client.experiments.create(
    project_id=project["id"],
    dataset_version_id="VERSION_UUID",
    target_column="label",
    problem_type="classification",
    engine="flaml",
    execution_mode="single"
)

# Enterprise: parallel engines
exp = client.experiments.create(
    project_id=project["id"],
    dataset_version_id="VERSION_UUID",
    target_column="label",
    problem_type="classification",
    execution_mode="parallel",
    engines=["h2o", "flaml"]
)
print(client.experiments.engine_runs(exp.get("experiment_id") or exp["id"]))

Links

• Platform
• API Documentation
• Website

License

Business Source License 1.1