feat: add Feast feature store integration

README.md

@@ -156,6 +156,40 @@ for version in client.list_model_versions("my-model"):
     print(f"Version: {version.name}")
 ```
 
+### Manage features with Feast
+
+**Install Feast support:**
+```bash
+pip install 'kubeflow[feast]'
+```
+
+Feast is a feature store that enables offline retrieval of historical datasets and online serving of features/data for ML applications.
+
+The FeastClient provides simplified initialization. Access full Feast functionality through the `feature_store` property:
+
+```python
+from kubeflow.feast import FeastClient
+
+# Initialize the Feast client with your feature repository path
+client = FeastClient(repo_path="/path/to/feast/repo")
+
+# Access full Feast functionality through feature_store property
+online_features = client.feature_store.get_online_features(
+    features=["feature_view:feature1", "feature_view:feature2"],
+    entity_rows=[{"entity_id": 1}, {"entity_id": 2}],
+)
+
+# Materialize features to the online store
+from datetime import datetime, timedelta
+end_date = datetime.now()
+start_date = end_date - timedelta(days=7)
+client.feature_store.materialize(start_date=start_date, end_date=end_date)
+
+# List all feature views
+for fv in client.feature_store.list_feature_views():
+    print(f"Feature view: {fv.name}")
+```
+
 ## Local Development
 
 Kubeflow Trainer client supports local development without needing a Kubernetes cluster.
@@ -186,6 +220,7 @@ job_id = client.train(trainer=CustomTrainer(func=train_fn))
 | **Kubeflow Trainer**        | ✅ **Available** | v2.0.0+         | Train and fine-tune AI models with various frameworks                 |
 | **Kubeflow Katib**          | ✅ **Available** | v0.19.0+        | Hyperparameter optimization                                           |
 | **Kubeflow Model Registry** | ✅ **Available** | v0.3.0+         | Manage model artifacts, versions and ML artifacts metadata            |
+| **Feast**                   | ✅ **Available** | v0.59.0+        | Feature store for offline and online feature serving                  |
 | **Kubeflow Pipelines**      | 🚧 Planned       | TBD             | Build, run, and track AI workflows                                    |
 | **Kubeflow Spark Operator** | 🚧 Planned       | TBD             | Manage Spark applications for data processing and feature engineering |
 

kubeflow/feast/__init__.py

@@ -0,0 +1,17 @@
+# Copyright 2025 The Kubeflow Authors.
+#
+# Licensed under the Apache License, Version 2.0 (the "License");
+# you may not use this file except in compliance with the License.
+# You may obtain a copy of the License at
+#
+#     http://www.apache.org/licenses/LICENSE-2.0
+#
+# Unless required by applicable law or agreed to in writing, software
+# distributed under the License is distributed on an "AS IS" BASIS,
+# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+# See the License for the specific language governing permissions and
+# limitations under the License.
+
+from kubeflow.feast.api.feast_client import FeastClient
+
+__all__ = ["FeastClient"]

kubeflow/feast/api/__init__.py

@@ -0,0 +1,13 @@
+# Copyright 2025 The Kubeflow Authors.
+#
+# Licensed under the Apache License, Version 2.0 (the "License");
+# you may not use this file except in compliance with the License.
+# You may obtain a copy of the License at
+#
+#     http://www.apache.org/licenses/LICENSE-2.0
+#
+# Unless required by applicable law or agreed to in writing, software
+# distributed under the License is distributed on an "AS IS" BASIS,
+# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+# See the License for the specific language governing permissions and
+# limitations under the License.

kubeflow/feast/api/feast_client.py

@@ -0,0 +1,101 @@
+# Copyright 2025 The Kubeflow Authors.
+#
+# Licensed under the Apache License, Version 2.0 (the "License");
+# you may not use this file except in compliance with the License.
+# You may obtain a copy of the License at
+#
+#     http://www.apache.org/licenses/LICENSE-2.0
+#
+# Unless required by applicable law or agreed to in writing, software
+# distributed under the License is distributed on an "AS IS" BASIS,
+# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+# See the License for the specific language governing permissions and
+# limitations under the License.
+
+from __future__ import annotations
+
+from typing import TYPE_CHECKING, Any
+
+if TYPE_CHECKING:
+    from feast import FeatureStore
+
+
+class FeastClient:
+    """Client for Feast feature store operations.
+
+    Feast is a feature store that enables offline retrieval of historical datasets
+    and online serving of features/data for ML applications.
+
+    This is a minimal wrapper that provides simplified initialization. For full Feast
+    functionality, use the `feature_store` property to access the underlying FeatureStore.
+
+    Requires the feast package to be installed. Install it with:
+
+        pip install 'kubeflow[feast]'
+
+    Example:
+        ```python
+        from kubeflow.feast import FeastClient
+
+        # Initialize client
+        client = FeastClient(repo_path="/path/to/feast/repo")
+
+        # Access full Feast functionality
+        client.feature_store.get_online_features(...)
+        client.feature_store.materialize(...)
+        ```
+    """
+
+    def __init__(self, repo_path: str | None = None, config: dict[str, Any] | None = None):
+        """Initialize the FeastClient.
+
+        Args:
+            repo_path: Path to the Feast repository. If not provided, uses the current directory.
+            config: Optional configuration dictionary for Feast FeatureStore.
+                   If provided, takes precedence over repo_path.
+
+        Raises:
+            ImportError: If feast is not installed.
+        """
+        try:
+            from feast import FeatureStore
+        except ImportError as e:
+            raise ImportError(
+                "feast is not installed. Install it with:\n\n"  # fmt: skip
+                "  pip install 'kubeflow[feast]'\n"
+            ) from e
+
+        if config is not None:
+            self._feature_store: FeatureStore = FeatureStore(config=config)
+        else:
+            self._feature_store: FeatureStore = FeatureStore(repo_path=repo_path)
+
+    @property
+    def feature_store(self) -> FeatureStore:
+        """Access the underlying Feast FeatureStore instance.
+
+        Use this property to access the full Feast API for operations like:
+        - get_online_features() / get_historical_features()
+        - materialize() / materialize_incremental()
+        - apply() - Deploy feature definitions
+        - list_feature_views() / list_entities() / list_data_sources()
+
+        Returns:
+            The Feast FeatureStore instance.
+
+        Example:
+            ```python
+            client = FeastClient(repo_path="/path/to/feast/repo")
+
+            # Get online features
+            features = client.feature_store.get_online_features(
+                features=["feature_view:feature1"],
+                entity_rows=[{"entity_id": 1}],
+            )
+
+            # List feature views
+            for fv in client.feature_store.list_feature_views():
+                print(fv.name)
+            ```
+        """
+        return self._feature_store