Added SQL fileto demostrate database set up

README.md

@@ -1,5 +1,5 @@
 ---
-title: Template Final Assignment
+title: GAIA Agent
 emoji: 🕵🏻‍♂️
 colorFrom: indigo
 colorTo: indigo
@@ -12,4 +12,12 @@ hf_oauth: true
 hf_oauth_expiration_minutes: 480
 ---
 
-Check out the configuration reference at https://huggingface.co/docs/hub/spaces-config-reference
+**Project Overview**
+Developed an Agentic RAG system using LangGraph that orchestrates a multi-step workflow combining retrieval and reasoning capabilities. The agent integrates multiple search tools (Wikipedia, Arxiv, web search via Tavily), mathematical operations, and a Supabase vector database for semantic similarity search and question retrieval. For databse setup, run *supabase_sql_setup.sql*
+
+**Evaluation Process**
+The project was evaluated using the GAIA benchmark, specifically testing against 20 questions extracted from the level 1 validation set. This rigorous evaluation measured the agent's ability to handle complex, multi-step reasoning tasks. Performance was assessed through automated scoring, providing detailed metrics including overall accuracy percentage and correct answer counts.
+
+
+Check out the configuration reference at https://huggingface.co/docs/hub/spaces-config-reference
+

supabase_sql_setup.sql

@@ -0,0 +1,41 @@
+-- First, make sure you have the vector extension enabled
+-- First, make sure you have the vector extension enabled
+CREATE EXTENSION IF NOT EXISTS vector;
+
+-- Create the documents table if it doesn't exist
+CREATE TABLE IF NOT EXISTS documents (
+    id SERIAL PRIMARY KEY,
+    content TEXT,
+    metadata JSONB,
+    embedding VECTOR(768)
+);
+
+-- Create the similarity search function that LangChain expects
+CREATE OR REPLACE FUNCTION match_documents_langchain_2(
+    query_embedding VECTOR(768),
+    match_threshold FLOAT DEFAULT 0.6,
+    match_count INT DEFAULT 10
+)
+RETURNS TABLE (
+    id BIGINT,
+    content TEXT,
+    metadata JSONB,
+    similarity FLOAT
+)
+LANGUAGE SQL STABLE
+AS $$
+    SELECT
+        documents.id,
+        documents.content,
+        documents.metadata,
+        1 - (documents.embedding <=> query_embedding) AS similarity
+    FROM documents
+    WHERE 1 - (documents.embedding <=> query_embedding) > match_threshold
+    ORDER BY documents.embedding <=> query_embedding
+    LIMIT match_count;
+$$;
+
+-- Create an index on the embedding column for better performance
+CREATE INDEX IF NOT EXISTS documents_embedding_idx 
+ON documents USING ivfflat (embedding vector_cosine_ops)
+WITH (lists = 100);