Move GenAI page to getting-started directory and update SUMMARY.md

Co-Authored-By: Francisco Javier Arceo <farceo@redhat.com>
feast-dev · franciscojavierarceo · Jun 2, 2025 · May 24, 2025 · May 24, 2025 · May 24, 2025
commit 06ae0f7e88ace2d722092bd9a1a85680d0e9b81a
diff --git a/docs/SUMMARY.md b/docs/SUMMARY.md
@@ -9,7 +9,7 @@
 ## Getting started
 
 * [Quickstart](getting-started/quickstart.md)
-* [GenAI](getting-started/genai.md)
+* [Feast for Generative AI](getting-started/genai.md)
 * [Architecture](getting-started/architecture/README.md)
   * [Overview](getting-started/architecture/overview.md)
   * [Language](getting-started/architecture/language.md)

diff --git a/docs/genai.md b/docs/genai.md
@@ -2,7 +2,7 @@
 
 ## Overview
 
-Feast provides robust support for Generative AI applications, enabling teams to build, deploy, and manage feature infrastructure for Large Language Models (LLMs) and other Generative AI (GenAI) applications. With Feast's vector database integrations and feature management capabilities, teams can implement production-ready Retrieval Augmented Generation (RAG) systems and other GenAI applications with the same reliability and operational excellence as traditional ML systems.
+Feast provides robust support for Generative AI applications, enabling teams to build, deploy, and manage feature infrastructure for Large Language Models (LLMs) and other generative AI systems. With Feast's vector database integrations and feature management capabilities, teams can implement production-ready Retrieval Augmented Generation (RAG) systems and other GenAI applications with the same reliability and operational excellence as traditional ML systems.
 
 ## Key Capabilities for GenAI
 
@@ -17,17 +17,17 @@ Feast integrates with popular vector databases to store and retrieve embedding v
 * **Qdrant**: Purpose-built vector database integration
 
 These integrations allow you to:
-- Store embeddings as features
+- Store document embeddings as features
 - Perform vector similarity search to find relevant context
 - Retrieve both vector embeddings and traditional features in a single API call
 
 ### Retrieval Augmented Generation (RAG)
 
 Feast simplifies building RAG applications by providing:
 
-1. **Embedding storage**: Store and version embeddings alongside your other features
-2. **Vector similarity search**: Find the most relevant data/documents for a given query
-3. **Feature retrieval**: Combine embeddings with structured features for richer context
+1. **Document embedding storage**: Store and version document embeddings alongside your other features
+2. **Vector similarity search**: Find the most relevant documents for a given query
+3. **Feature retrieval**: Combine document embeddings with structured features for richer context
 4. **Versioning and governance**: Track changes to your document repository over time
 
 The typical RAG workflow with Feast involves:
@@ -59,12 +59,105 @@ The transformation workflow typically involves:
 
 ### Feature Transformation for LLMs
 
-Feast supports transformations that can be used to:
+Feast supports on-demand transformations that can be used to:
 
 * Process raw text into embeddings
 * Chunk documents for more effective retrieval
 * Normalize and preprocess features before serving to LLMs
-* Apply custom transformations to adapt features for specific LLM requirements
+
+### Getting Started with Feast for GenAI
+
+#### Installation
+
+To use Feast with vector database support, install with the appropriate extras:
+
+```bash
+# For Milvus support
+pip install feast[milvus,nlp]
+
+# For Elasticsearch support
+pip install feast[elasticsearch]
+
+# For Qdrant support
+pip install feast[qdrant]
+
+# For SQLite support (Python 3.10 only)
+pip install feast[sqlite_vec]
+```
+
+#### Configuration
+
+Configure your feature store to use a vector database as the online store:
+
+```yaml
+project: genai-project
+provider: local
+registry: data/registry.db
+online_store:
+  type: milvus
+  path: data/online_store.db
+  vector_enabled: true
+  embedding_dim: 384  # Adjust based on your embedding model
+  index_type: "IVF_FLAT"
+
+offline_store:
+  type: file
+entity_key_serialization_version: 3
+```
+
+#### Defining Vector Features
+
+Create feature views with vector index support:
+
+```python
+from feast import FeatureView, Field, Entity
+from feast.types import Array, Float32, String
+
+document = Entity(
+    name="document_id",
+    description="Document identifier",
+    join_keys=["document_id"],
+)
+
+document_embeddings = FeatureView(
+    name="document_embeddings",
+    entities=[document],
+    schema=[
+        Field(
+            name="vector",
+            dtype=Array(Float32),
+            vector_index=True,  # Enable vector search
+            vector_search_metric="COSINE",  # Similarity metric
+        ),
+        Field(name="document_id", dtype=String),
+        Field(name="content", dtype=String),
+    ],
+    source=document_source,
+    ttl=timedelta(days=30),
+)
+```
+
+#### Retrieving Similar Documents
+
+Use the `retrieve_online_documents_v2` method to find similar documents:
+
+```python
+# Generate query embedding
+query = "How does Feast support vector databases?"
+query_embedding = embed_text(query)  # Your embedding function
+
+# Retrieve similar documents
+context_data = store.retrieve_online_documents_v2(
+    features=[
+        "document_embeddings:vector",
+        "document_embeddings:document_id",
+        "document_embeddings:content",
+    ],
+    query=query_embedding,
+    top_k=3,
+    distance_metric='COSINE',
+).to_df()
+```
 
 ## Use Cases