Convex MCP server

--- name: Convex title: Fivetran for Convex | Configuration and documentation description: Connect data sources to Convex with our partner-built Convex destination connector. Explore documentation and start syncing your applications, databases, events, files, and more. menuPosition: 55 hidden: true --- # Convex {% badge text="Partner-Built" /%} {% availabilityBadge connector="convex_destination" /%} [Convex](https://convex.dev) is a full-stack TypeScript development platform with product-centric APIs. It can replace your database and server functions. > NOTE: Fivetran supports Convex as both a partner-built [database connector](/docs/databases/convex) and a partner-built destination. > NOTE: This destination is [partner-built](/docs/partner-built-program). For any questions related to the Convex destination and its documentation, refer to Convex's support team. For SLA details, see [Convex's Status and Guarantees documentation](https://docs.convex.dev/production/state). --- ## Setup guide Follow our [step-by-step Convex setup guide](/docs/destinations/convex/setup-guide) to connect Convex as a destination with Fivetran. --- ## Schema information Fivetran tries to replicate the database and columns from your data source to your Convex destination according to Fivetran's [standard database update strategies](/docs/databases#transformationandmappingoverview). Once Fivetran connects to your Convex destination, the connector will attempt to load your data. It may ask you to update your `convex/schema.ts` in your destination to match the format of your source. Once the `convex/schema.ts` matches the source format, data will continue to sync. ### Type transformation mapping The Convex destination extracts data from your source, and it matches supported [Fivetran data types](/docs/destinations#datatypes) to [Convex data types](https://docs.convex.dev/database/types). We use the following data type conversions: | Fivetran Type | Convex Type | Equivalence | | ------------- | ----------- | ----------- | | BOOLEAN | Boolean | Exact | | SHORT | Float64 | Inexact | | INT | Float64 | Inexact | | LONG | Int64 | Exact | | DECIMAL | String | Inexact | | FLOAT | Float64 | Inexact | | DOUBLE | Float64 | Exact | | NAIVEDATE | String | Inexact | | NAIVEDATETIME | String | Inexact | | UTCDATETIME | Float64 | Inexact | | BINARY | Bytes | Exact | | STRING | String | Exact | | NULL | Null | Exact | | JSON | Object | Inexact | > NOTE: Short/Int are converted to float64 for ease of use in javascript (as `number`). There is no data loss as `Number.MAX_SAFE_INTEGER = 2^53 - 1`. Decimal is converted to STRING to ensure no data loss (e.g. "1234.5678"). > NOTE: Naive date uses standard string representation of `YYYY-MM-DD`. Naive datetime uses standard string representation of `YYYY-MM-DD HH:MM:SS`. > NOTE: UTC datetime uses milliseconds since UNIX epoch. ### Fivetran-generated data Fivetran adds a single `fivetran` column containing a Convex object to the source data. Some of the columns (`synced`, `deleted`) are for internal purposes. ### Table and column name transformations If your source table is `default.cars`, then your Convex table will be named `default_cars`. Convex deployments do not have a concept of namespaced tables, so it uses this notation to preserve the namespace information. Column names that begin with `_` are not supported in Convex. Instead, those columns are synced to the destination nested within the `fivetran.columns` column. For example, a column named `_line` would be synced as a nested column named `fivetran.columns.line`.