[Store][Supabase] Add store

docs/components/store.rst

@@ -48,6 +48,7 @@ You can find more advanced usage in combination with an Agent using the store fo
 * `Similarity Search with Symfony Cache (RAG)`_
 * `Similarity Search with Typesense (RAG)`_
 * `Similarity Search with Weaviate (RAG)`_
+* `Similarity Search with Supabase (RAG)`_
 
 .. note::
 
@@ -70,6 +71,7 @@ Supported Stores
 * `Pinecone`_ (requires `probots-io/pinecone-php` as additional dependency)
 * `Postgres`_ (requires `ext-pdo`)
 * `Qdrant`_
+* `Supabase`_ (requires manual database setup)
 * `SurrealDB`_
 * `Symfony Cache`_ (requires `symfony/cache` as additional dependency)
 * `Typesense`_
@@ -137,6 +139,7 @@ This leads to a store implementing two methods::
 .. _`Similarity Search with Qdrant (RAG)`: https://github.com/symfony/ai/blob/main/examples/rag/qdrant.php
 .. _`Similarity Search with SurrealDB (RAG)`: https://github.com/symfony/ai/blob/main/examples/rag/surrealdb.php
 .. _`Similarity Search with Typesense (RAG)`: https://github.com/symfony/ai/blob/main/examples/rag/typesense.php
+.. _`Similarity Search with Supabase (RAG)`: https://github.com/symfony/ai/blob/main/examples/rag/supabase.php
 .. _`Similarity Search with Weaviate (RAG)`: https://github.com/symfony/ai/blob/main/examples/rag/weaviate.php
 .. _`Azure AI Search`: https://azure.microsoft.com/products/ai-services/ai-search
 .. _`Chroma`: https://www.trychroma.com/
@@ -154,3 +157,4 @@ This leads to a store implementing two methods::
 .. _`Typesense`: https://typesense.org/
 .. _`Symfony Cache`: https://symfony.com/doc/current/components/cache.html
 .. _`Weaviate`: https://weaviate.io/
+.. _`Supabase`: https://https://supabase.com/

docs/components/store/supabase.rst

@@ -0,0 +1,226 @@
+Supabase Bridge
+===============
+
+The Supabase bridge provides vector storage capabilities using `pgvector`_ extension through the REST API.
+
+.. note::
+
+Unlike the Postgres Store, the Supabase Store requires manual setup of the database schema because Supabase doesn't
+allow arbitrary SQL execution via REST API.
+
+Requirements
+~~~~~~~~~~~~
+
+* Enable `pgvector extension`_ in the relevant schema of your Supabase project for using `vector`_ column types.
+* Add columns for embedding (type `vector`) and metadata (type `jsonb`) to your table
+* Pre-configured RPC `function`_ for similarity search
+
+See section below for detailed SQL commands.
+
+Database Setup
+--------------
+
+Execute the following SQL commands in your Supabase SQL Editor:
+
+Enable ``pgvector`` extension
+~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
+
+.. code-block:: sql
+
+    CREATE EXTENSION IF NOT EXISTS vector;
+
+Create the `documents` table
+~~~~~~~~~~~~~~~~~~~~~~~~~~~~
+
+.. code-block:: sql
+
+    CREATE TABLE IF NOT EXISTS documents (
+        id UUID PRIMARY KEY DEFAULT gen_random_uuid(),
+        embedding vector(768) NOT NULL,
+        metadata JSONB
+    );
+
+Create the similarity search function
+~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
+
+.. code-block:: sql
+
+    CREATE OR REPLACE FUNCTION match_documents(
+        query_embedding vector(768),
+        match_count int DEFAULT 10,
+        match_threshold float DEFAULT 0.0
+    )
+    RETURNS TABLE (
+        id UUID,
+        embedding vector,
+        metadata JSONB,
+        score float
+    )
+    LANGUAGE sql
+    AS $$
+        SELECT
+            documents.id,
+            documents.embedding,
+            documents.metadata,
+            1- (documents.embedding <=> query_embedding) AS score
+        FROM documents
+        WHERE 1- (documents.embedding <=> query_embedding) >= match_threshold
+        ORDER BY documents.embedding <=> query_embedding ASC
+        LIMIT match_count;
+    $$;
+
+Create an index for better performance
+~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
+
+.. code-block:: sql
+
+    CREATE INDEX IF NOT EXISTS documents_embedding_idx
+    ON documents USING ivfflat (embedding vector_cosine_ops);
+
+Configuration
+-------------
+
+Basic Configuration
+~~~~~~~~~~~~~~~~~~~
+
+.. code-block:: php
+
+    use Symfony\AI\Store\Bridge\Supabase\Store;
+    use Symfony\Component\HttpClient\HttpClient;
+
+    $store = new Store(
+        HttpClient::create(),
+        'https://your-project.supabase.co',
+        'your-anon-key',
+        'documents',        // table name
+        'embedding',        // vector field name
+        768,               // vector dimension (depending on your embedding model)
+        'match_documents'   // function name
+    );
+
+Bundle Configuration
+~~~~~~~~~~~~~~~~~~~~
+
+.. code-block:: yaml
+
+    # config/packages/ai.yaml
+    ai:
+        store:
+            supabase:
+                my_supabase_store:
+                    url: 'https://your-project.supabase.co'
+                    api_key: '%env(SUPABASE_API_KEY)%'
+                    table: 'documents'
+                    vector_field: 'embedding'
+                    vector_dimension: 768
+                    function_name: 'match_documents'
+
+Environment Variables
+~~~~~~~~~~~~~~~~~~~~~
+
+.. code-block:: bash
+
+    # .env.local
+    SUPABASE_URL=https://your-project.supabase.co
+    SUPABASE_API_KEY=your-supabase-anon-key
+
+Usage
+-----
+
+Adding Documents
+~~~~~~~~~~~~~~~~
+
+.. code-block:: php
+
+    use Symfony\AI\Platform\Vector\Vector;
+    use Symfony\AI\Store\Document\Metadata;
+    use Symfony\AI\Store\Document\VectorDocument;
+    use Symfony\Component\Uid\Uuid;
+
+    $document = new VectorDocument(
+        Uuid::v4(),
+        new Vector([0.1, 0.2, 0.3, /* ... 768 dimensions */]),
+        new Metadata(['title' => 'My Document', 'category' => 'example'])
+    );
+
+    $store->add($document);
+
+Querying Documents
+~~~~~~~~~~~~~~~~~~
+
+.. code-block:: php
+
+    $queryVector = new Vector([0.1, 0.2, 0.3, /* ... 768 dimensions */]);
+
+    $results = $store->query($queryVector, [
+        'max_items' => 10,
+        'min_score' => 0.7
+    ]);
+
+    foreach ($results as $document) {
+        echo "ID: " . $document->id . "\n";
+        echo "Score: " . $document->score . "\n";
+        echo "Metadata: " . json_encode($document->metadata->getArrayCopy()) . "\n";
+    }
+
+Customization
+-------------
+
+You can customize the Supabase setup for different requirements:
+
+Table Name
+~~~~~~~~~~
+
+Change ``documents`` to your preferred table name in both the SQL setup and configuration.
+
+Vector Field Name
+~~~~~~~~~~~~~~~~~
+
+Change ``embedding`` to your preferred field name in both the SQL setup and configuration.
+
+Vector Dimension
+~~~~~~~~~~~~~~~~
+
+Change ``768`` to match your embedding model's dimensions in both the SQL setup and configuration.
+
+Distance Metric
+~~~~~~~~~~~~~~~
+
+* Cosine: ``<=>`` (default, recommended for most embeddings)
+* Euclidean: ``<->``
+* Inner Product: ``<#>``
+
+Index Type
+~~~~~~~~~~
+
+* ``ivfflat``: Good balance of speed and accuracy
+* ``hnsw``: Better for high-dimensional vectors (requires PostgreSQL 14+)
+
+Limitations
+-----------
+
+* Manual schema setup required (no automatic table creation)
+* Limited to Supabase's REST API capabilities
+* Requires pre-configured RPC functions for complex queries
+* Vector dimension must be consistent across all documents
+
+Performance Considerations
+--------------------------
+
+* Use appropriate index types based on your vector dimensions
+* Consider using ``hnsw`` indexes for high-dimensional vectors
+* Batch document insertions when possible (up to 200 documents per request)
+* Monitor your Supabase usage limits and quotas
+
+Security Considerations
+-----------------------
+
+* Use row-level security (RLS) policies if needed
+* Consider using service role keys for server-side operations
+* Validate vector dimensions in your application code
+* Implement proper error handling for API failures
+
+.. _`pgvector`: https://github.com/pgvector/pgvector
+.. _`pgvector extension`: https://supabase.com/docs/guides/database/extensions/pgvector
+.. _`vector`: https://supabase.com/docs/guides/ai/vector-columns
+.. _`function`: https://supabase.com/docs/guides/database/functions

examples/.env

@@ -156,3 +156,11 @@ SCALEWAY_SECRET_KEY=
 
 # For using DeepSeek
 DEEPSEEK_API_KEY=
+
+# Supabase (store)
+SUPABASE_URL=
+SUPABASE_API_KEY=
+SUPABASE_TABLE=documents
+SUPABASE_VECTOR_FIELD=embedding
+SUPABASE_VECTOR_DIMENSION=768 # when using Ollama with nomic-embed-text
+SUPABASE_MATCH_FUNCTION=match_documents

examples/rag/supabase.php

@@ -0,0 +1,69 @@
+<?php
+
+/*
+ * This file is part of the Symfony package.
+ *
+ * (c) Fabien Potencier <[email protected]>
+ *
+ * For the full copyright and license information, please view the LICENSE
+ * file that was distributed with this source code.
+ */
+
+use Symfony\AI\Agent\Agent;
+use Symfony\AI\Agent\Toolbox\AgentProcessor;
+use Symfony\AI\Agent\Toolbox\Tool\SimilaritySearch;
+use Symfony\AI\Agent\Toolbox\Toolbox;
+use Symfony\AI\Fixtures\Movies;
+use Symfony\AI\Platform\Bridge\Ollama\PlatformFactory;
+use Symfony\AI\Platform\Message\Message;
+use Symfony\AI\Platform\Message\MessageBag;
+use Symfony\AI\Store\Bridge\Supabase\Store;
+use Symfony\AI\Store\Document\Loader\InMemoryLoader;
+use Symfony\AI\Store\Document\Metadata;
+use Symfony\AI\Store\Document\TextDocument;
+use Symfony\AI\Store\Document\Vectorizer;
+use Symfony\AI\Store\Indexer;
+use Symfony\Component\Uid\Uuid;
+
+require_once dirname(__DIR__).'/bootstrap.php';
+
+$store = new Store(
+    httpClient: http_client(),
+    url: env('SUPABASE_URL'),
+    apiKey: env('SUPABASE_API_KEY'),
+    table: env('SUPABASE_TABLE'),
+    vectorFieldName: env('SUPABASE_VECTOR_FIELD'),
+    vectorDimension: (int) env('SUPABASE_VECTOR_DIMENSION'),
+    functionName: env('SUPABASE_MATCH_FUNCTION'),
+);
+
+$documents = [];
+
+foreach (Movies::all() as $movie) {
+    $documents[] = new TextDocument(
+        id: Uuid::v4(),
+        content: 'Title: '.$movie['title'].\PHP_EOL.'Director: '.$movie['director'].\PHP_EOL.'Description: '.$movie['description'],
+        metadata: new Metadata($movie),
+    );
+}
+
+$platform = PlatformFactory::create(env('OLLAMA_HOST_URL'), http_client());
+
+$vectorizer = new Vectorizer($platform, env('OLLAMA_EMBEDDINGS'));
+$loader = new InMemoryLoader($documents);
+$indexer = new Indexer($loader, $vectorizer, $store, logger: logger());
+$indexer->index();
+
+$similaritySearch = new SimilaritySearch($vectorizer, $store);
+$toolbox = new Toolbox([$similaritySearch], logger: logger());
+$processor = new AgentProcessor($toolbox);
+$agent = new Agent($platform, env('OLLAMA_LLM'), [$processor], [$processor]);
+
+$messages = new MessageBag(
+    Message::forSystem('Please answer all user questions only using SimilaritySearch function.'),
+    Message::ofUser('Which movie fits the theme of technology?')
+);
+
+$result = $agent->call($messages);
+
+echo $result->getContent().\PHP_EOL;

src/ai-bundle/config/options.php

@@ -647,6 +647,24 @@
                             ->end()
                         ->end()
                     ->end()
+                    ->arrayNode('supabase')
+                        ->useAttributeAsKey('name')
+                        ->arrayPrototype()
+                            ->children()
+                                ->stringNode('http_client')
+                                    ->cannotBeEmpty()
+                                    ->defaultValue('http_client')
+                                    ->info('Service ID of the HTTP client to use')
+                                ->end()
+                                ->stringNode('url')->isRequired()->cannotBeEmpty()->end()
+                                ->stringNode('api_key')->isRequired()->cannotBeEmpty()->end()
+                                ->stringNode('table')->end()
+                                ->stringNode('vector_field')->end()
+                                ->integerNode('vector_dimension')->end()
+                                ->stringNode('function_name')->end()
+                            ->end()
+                        ->end()
+                    ->end()
                     ->arrayNode('typesense')
                         ->useAttributeAsKey('name')
                         ->arrayPrototype()