メニュー
Train and deploy propensity models using Databricks AutoML. Use when the agent needs campaign-specific propensity scores and no Model Serving endpoint exists yet, or when retraining is needed for a new campaign type.
Codex または Claude でインストール この Prompt をコピーして Codex、Claude、または他のアシスタントに貼り付けると、Skill ページを確認してインストールできます。
Best practices for cross-campaign CDP disposition reasoning. Use when evaluating customer engagement decisions involving frequency caps, campaign prioritization, channel selection, and send-time optimization.
DBOS Python SDK for building reliable, fault-tolerant applications with durable workflows. Use this skill when writing Python code with DBOS, creating workflows and steps, using queues, using DBOSClient from external applications, or building applications that need to be resilient to failures.
A simple greeting skill that says hello to a user by name. Use when the user asks for a greeting or hello message.
SOC 職業分類に基づく
| name | automl-propensity |
| description | Train and deploy propensity models using Databricks AutoML. Use when the agent needs campaign-specific propensity scores and no Model Serving endpoint exists yet, or when retraining is needed for a new campaign type. |
Spawn a Databricks AutoML workflow to train a propensity model for a specific campaign type (e.g., cart_recovery, seasonal_purchase, loyalty_renewal). The trained model is registered in Unity Catalog and deployed to a Model Serving endpoint.
maestro_cdp.models.propensity_{campaign_type}score_propensity tool to calltrain_propensity.pyTrain a propensity model for a given campaign type.
Arguments:
--campaign-type (required): The campaign type to train for (e.g., cart_recovery)--training-table (optional): UC table with training data (default: maestro_cdp.features.customer_events)--target-column (optional): Target column name (default: converted)--profile (optional): Databricks CLI profile (default: 9cefok)Output: JSON with model_name, endpoint_name, experiment_url, accuracy
score_propensity returns cached synthetic scores (0.81 for Cindy). This skill is discoverable but not invoked.score_propensity calls it instead of returning cached data.