RAG Custom Embeddings

This cookbook is an implementation of Agentic RAG using Sentence Transformer Reranker with multilingual data.

rag_custom_embeddings.py

"""
Rag Custom Embeddings
=============================

This cookbook is an implementation of Agentic RAG using Sentence Transformer Reranker with multilingual data.
"""

from agno.agent import Agent
from agno.knowledge.embedder.sentence_transformer import SentenceTransformerEmbedder
from agno.knowledge.knowledge import Knowledge
from agno.knowledge.reranker.sentence_transformer import SentenceTransformerReranker
from agno.models.openai import OpenAIResponses
from agno.vectordb.pgvector import PgVector

search_results = [
    "Organic skincare for sensitive skin with aloe vera and chamomile.",
    "New makeup trends focus on bold colors and innovative techniques",
    "Bio-Hautpflege für empfindliche Haut mit Aloe Vera und Kamille",
    "Neue Make-up-Trends setzen auf kräftige Farben und innovative Techniken",
    "Cuidado de la piel orgánico para piel sensible con aloe vera y manzanilla",
    "Las nuevas tendencias de maquillaje se centran en colores vivos y técnicas innovadoras",
    "针对敏感肌专门设计的天然有机护肤产品",
    "新的化妆趋势注重鲜艳的颜色和创新的技巧",
    "敏感肌のために特別に設計された天然有機スキンケア製品",
    "新しいメイクのトレンドは鮮やかな色と革新的な技術に焦点を当てています",
]

knowledge = Knowledge(
    vector_db=PgVector(
        db_url="postgresql+psycopg://ai:ai@localhost:5532/ai",
        table_name="sentence_transformer_rerank_docs",
        embedder=SentenceTransformerEmbedder(
            id="sentence-transformers/paraphrase-multilingual-MiniLM-L12-v2"
        ),
        reranker=SentenceTransformerReranker(model="BAAI/bge-reranker-v2-m3"),
    ),
)

for result in search_results:
    knowledge.insert(
        text_content=result,
        metadata={
            "source": "search_results",
        },
    )


# ---------------------------------------------------------------------------
# Create Agent
# ---------------------------------------------------------------------------
agent = Agent(
    model=OpenAIResponses(id="gpt-5.2"),
    knowledge=knowledge,
    search_knowledge=True,
    instructions=[
        "Include sources in your response.",
        "Always search your knowledge before answering the question.",
    ],
    markdown=True,
)


# ---------------------------------------------------------------------------
# Run Agent
# ---------------------------------------------------------------------------
if __name__ == "__main__":
    test_queries = [
        "What organic skincare products are good for sensitive skin?",
        "Tell me about makeup trends in different languages",
        "Compare skincare and makeup information across languages",
    ]

    for query in test_queries:
        agent.print_response(
            query,
            stream=True,
            show_full_reasoning=True,
        )

Run the Example

1. Set up your virtual environment

   Mac

   uv venv --python 3.12
   source .venv/bin/activate
   

   uv venv --python 3.12
   .venv\Scripts\activate
   
2. Install dependencies

   uv pip install -U agno numpy openai pgvector psycopg-binary sentence-transformers sqlalchemy
   
3. Export your OpenAI API key

   Mac/Linux

   export OPENAI_API_KEY="your_openai_api_key_here"
   

   $Env:OPENAI_API_KEY="your_openai_api_key_here"
   
4. Run PgVector

   docker run -d \
     -e POSTGRES_DB=ai \
     -e POSTGRES_USER=ai \
     -e POSTGRES_PASSWORD=ai \
     -e PGDATA=/var/lib/postgresql/data/pgdata \
     -v pgvolume:/var/lib/postgresql/data \
     -p 5532:5432 \
     --name pgvector \
     agnohq/pgvector:18
   
5. Run the example

   Save the code above as rag_custom_embeddings.py, then run:

   python rag_custom_embeddings.py
   

Full source: cookbook/02_agents/07_knowledge/rag_custom_embeddings.py