The SupabaseSink class provides the capabilities to interface with Supabase databases for the purpose of storing and querying vectorized data, which is essential for operations like semantic searches.


Required properties:

  • database_connection: The connection string necessary to connect to the Supabase database.
  • collection_name: The name of the collection within the database where the vectors will be stored.
from neumai.SinkConnectors import SupabaseSink

# Initialize the SupabaseSink connector with the database connection string
supabase_sink = SupabaseSink(
    database_connection = "postgresql://postgres:<password>@<host>:<port>/postgres",
    collection_name = "collection-name"