Connectors

Read And Write Every Data Source

Zingg reads from and writes to 24+ data platforms, databases, stores and file formats, plus anything with a Spark connector. Every connector is both a source and a sink.

Data platforms Cloud storage Relational databases NoSQL, search and graph File formats Any Spark connector

Data platforms

Warehouses And Lakehouses

Zingg runs where your tables already live, reading them and writing resolved entities back beside them.

Databricks

Delta tables and Unity Catalog, read and written on Databricks Spark.

Learn more Docs

Snowflake

Snowflake tables through the Snowflake Spark connector, or natively inside Snowflake with Zingg Enterprise.

Learn more Docs

Microsoft Fabric

Lakehouse tables and OneLake files from Fabric Spark.

Learn more Docs

Google BigQuery

BigQuery tables through the BigQuery Spark connector, with Cloud Storage for working files.

Learn more Docs

Exasol

The Exasol in-memory analytics database through the Exasol Spark connector.

Docs

Cloud storage

Object Storage

Point Zingg at a bucket for input files, output and the Zingg model directory.

Amazon S3

Files in S3 buckets, from AWS Glue, EMR or anywhere Spark runs.

Learn more Docs

Azure Storage

Azure Blob Storage and Data Lake Storage when Zingg runs on Azure.

Docs

Google Cloud Storage

Cloud Storage buckets through the Hadoop GCS connector.

Learn more Docs

Relational databases

Any Database With A JDBC Driver

Add the database’s JDBC driver to Zingg’s runtime properties and read or write any table.

PostgreSQL

Nodes and relationships through the Neo4j Spark connector, for an entity knowledge graph.

Docs

File formats

XLSX workbooks through the Spark Excel connector.

Docs

Zingg Pipes

Anything With A Spark Connector

A pipe is a name, a Spark connector format and the properties that connector needs. Set pipes for the data Zingg reads and the output it writes, and mix them freely: read a CRM in PostgreSQL, write golden records to Snowflake.

Several input pipes in one run
Input and output on different systems
Spark read and write options passed through

Read about Zingg Pipes

config.jsonPostgreSQL in, Snowflake out

{
  "data": [{
    "name": "crm_customers",
    "format": "jdbc",
    "props": {
      "url": "jdbc:postgresql://localhost:5432/crm",
      "dbtable": "customers",
      "driver": "org.postgresql.Driver",
      "user": "zingg"
    }
  }],
  "output": [{
    "name": "resolved_customers",
    "format": "net.snowflake.spark.snowflake",
    "props": {
      "sfDatabase": "MDM",
      "sfSchema": "PUBLIC",
      "dbtable": "RESOLVED_CUSTOMERS"
    },
    "mode": "Append"
  }]
}

FAQ

Connector Questions

Which data sources does Zingg support?

Databricks, Snowflake, Microsoft Fabric, BigQuery, Exasol, Amazon S3, Azure Storage, Google Cloud Storage, any database with a JDBC driver such as PostgreSQL, MySQL, SQL Server and Oracle, MongoDB, Cassandra, Elasticsearch, Neo4j, and files in Parquet, Delta, CSV, JSON, Avro and Excel. Anything else with a Spark connector works through a Zingg Pipe.

Is every connector both a source and a sink?

Yes. Zingg reads its input through a pipe and writes its output through a pipe, and any connector can be used for either.

Can Zingg read from one system and write to another?

Yes. Input and output are configured separately, and the input can combine several pipes, so you can match a CRM table against files in S3 and write the result to your warehouse.

What is a Zingg Pipe?

A name, a Spark connector format and the properties that connector needs. Pipes are set in the Zingg configuration for the data Zingg reads and the output it writes.

Does my data leave my platform?

No. Zingg runs on your Spark or Snowflake compute and reads and writes through your own connectors and credentials.

Resolve Entities Where Your Data Lives

Connect Zingg to your sources and write golden records back to your platform.

Get Zingg Enterprise Connector Docs