artie

Keep warehouse data current without building or operating a CDC stack. Artie MCP lets you set up and operate continuous replication from a conversation. Tell Claude which database, tables, and destination you need. Artie creates and runs the pipeline that keeps the destination current as production rows change. Use it when you need fresh operational data in Snowflake, BigQuery, Databricks, Redshift, ClickHouse, or another supported destination, without wiring together connectors, replication jobs, backfills, and schema-change handling yourself. What you can do - Create a replication pipeline from a source database to a destination - Select the tables to replicate, then add or remove tables later - Start, pause, resume, update, or delete pipelines - Run or cancel backfills - Detect schema changes and update pipeline configuration Example prompts - “Replicate `users`, `orders`, and `payments` from our Postgres primary to Snowflake.” - “Create a MongoDB to BigQuery pipeline for the `events` collection.” - “Add the `subscriptions` table to our existing MySQL pipeline.” - “Pause the staging pipeline while we upgrade the database.” - “Backfill the last six months of `orders` into Redshift.” - “Set up an SSH tunnel for this private Postgres instance.” Supported sources include Postgres, MySQL, MongoDB, SQL Server, DynamoDB, and Oracle. Supported destinations include Snowflake, BigQuery, Databricks, Redshift, and ClickHouse.