Quick Start

Quick Start

In this guide, we will create a Turbine project and get it up and running in a few minutes. To follow this guide, you'll need a Turbine API key. You can get one by signing up for a free account (opens in a new tab).

Install the Turbine SDK

pip install turbine-sdk

Create a Turbine project

To create a project, you have to provide details of

  • The data source you want to index. Currently, we support Postgres and Elasticsearch.
  • The vector database you want to use. Currently, we support Pinecone and Milvus.
  • The embedding model you want to use. Currently, we support OpenAI's text-embedding-ada-002, and all-MiniLM-L6-v2.

For example, to create a project for indexing data from a Postgres database, using text-embedding-ada-002 for creating embeddings, and Pinecone for storing the embeddings, you can use the following code:

from turbine import Turbine, ProjectConfig, DataSource, PostgresConfig
 
turbine = Turbine("your-api-key")
 
project_id = turbine.create_project(
    ProjectConfig(
        data_source=DataSource(
            type="postgres",
            config=PostgresConfig(
                url="postgres://user:pass@hostname:5432/postgres",
                table="table-name",
            ),
            fields=["column-1", "column-2"],
        ),
        vector_db="pinecone",
        embedding_model="text-embedding-ada-002",
    )
)

Learn more about project configuration.

Run search queries

Once you have created a project, you can start running search queries.

from turbine import Turbine
 
turbine = Turbine(api_key="your-api-key")
results = turbine.search("project-id", "your search query")
IntroductionProject Configuration