Skill UI

This guide outlines the complete MLOps workflow on Databricks. It covers feature engineering using Feature Store, tracking experiments with MLflow, managing model versions in the Model Registry, and deploying real-time inference endpoints via Mosaic AI Model Serving. Ideal for building robust, production-grade ML pipelines.

Mastering the full lifecycle of machine learning models from data ingestion to deployment. This expertise covers building robust, production-grade MLOps pipelines using orchestrators like Kubeflow, Airflow, or Prefect. Key skills include feature store implementation with Feast, rigorous data validation, experiment tracking (e.g., MLflow), model registry management, and automated deployment workflows. Essential for automating model training and ensuring reproducibility at scale.

Automates guidance for model registry manager tasks within ML deployment, covering serving, MLOps pipelines, monitoring, and production optimization whenever that trigger phrase appears.

Enables Claude to manage AI/ML model versions through the model-versioning-tracker plugin, automating logging, lineage tracking, performance summaries, and version retrieval whenever users request registry or deployment support.

Weights & Biases (W&B) automates experiment logging, real-time dashboards, hyperparameter sweeps, artifact tracking, and model registry/versioning for ML teams using PyTorch, TensorFlow, or HuggingFace workflows.

This guide explains how to register custom model slugs (variants) within the Pi Agent's configuration files. It is essential for ensuring that specialized OpenRouter variants or brand-new provider models, which are not part of the static bundled registry, resolve correctly and are available for accurate agent model invocation.

This skill helps developers programmatically discover, filter, and inspect the entire OpenMed model registry. It allows users to find the correct model key for specific clinical tasks (e.g., diseases, oncology, PII detection), filter by language, size, or task, and review model metadata before loading. It pairs with `loading-openmed-models`.

This skill manages the loading and efficient caching of OpenMed clinical/biomedical Named Entity Recognition (NER) models. Models are downloaded once from the Hugging Face Hub, enabling full offline operation afterward. It supports model reuse across multiple calls, configuring cache directories, specifying devices (CPU/GPU), and handling various model identifiers (registry keys, full HF IDs, or local paths).

Guides users to select a local, on-device Personally Identifiable Information (PII) detection model from a committed registry. It filters models based on language, required runtime environment (e.g., PyTorch, MLX), and size constraints. The process emphasizes building an offline shortlist and mandates rigorous pre-deployment recall validation and performance benchmarking to ensure secure, privacy-preserving deployment.

Establish model registry standards, governance controls, metadata schemas, approvals, and lifecycle policies for enterprise AI deployments. Includes MLflow setup, required metadata schema, and registration scripts for traceability and policy-driven promotion.