Quickstart

This walkthrough takes you from an empty workspace to a running pipeline. You will connect Nexa to a source system, pull in a data source, map it through the Curated and Consumption layers, and run the pipeline Nexa generates. Each step links to the deeper reference page for that screen. Budget 20–30 minutes for a first run against a small source.

Note

This assumes Nexa is already installed and you can sign in to the web UI. If you are standing the platform up, start with Prerequisites and Installation instead.

Before you begin

You need:

• Access to the Nexa web UI and a workspace to work in.
• Credentials for one source system — a PostgreSQL database, a Kafka-compatible message queue, or file storage (S3 / Unity Catalog volume).
• Your data platform (Snowflake or Databricks) already connected during install, so generated pipelines have somewhere to run.

Step 1 — Create a connection

A connection stores the reachable endpoint and credentials for one source system. Connections are managed under Settings → Connection Settings.

1. Open Settings and select the Connection Settings tab.
2. Add a connection for your source type:
   • RDS — host, port, database, and credentials for a relational database.
   • Message Queue — bootstrap servers and SASL/SCRAM authentication for Kafka.
   • File System — the path to your file storage.
3. Use Test connection to confirm Nexa can reach the endpoint before saving.

Nexa stores the secret securely rather than in plain config. See Connections overview, and the type-specific pages for PostgreSQL, Kafka, and S3 / volumes.

Step 2 — Create a connector

A connector binds a connection to the specific tables or files you want to ingest, and to a sync schedule. Connectors are managed on the Connectors screen.

1. Open Connectors and create a new connector.
2. Choose the connection from Step 1.
3. Select the tables (or files) to ingest.
4. Pick a sync mode — live for continuous sync or batch for a scheduled pull — and set a cron schedule for batch.

See Creating a connector and Execution modes for the full field reference.

Step 3 — Add the data source

With ingestion configured, the incoming data appears as a data source in Data Flow. This is where you confirm what Nexa detected and start modeling it.

1. Open Data Flow. Ingested sources show under the New Dataflows tab; already-modeled ones under Existing Dataflows.
2. Open the new source to review the tables and fields Nexa detected.
3. If you are working from a file instead of a live connector, you can also upload a CSV or JSON directly from the add-source flow.

See Adding a data source and the Data integration overview.

Step 4 — Map to the Curated layer

The Curated layer (silver) is where raw source fields become clean, conformed business entities. Nexa proposes field mappings and PII tags with AI assistance; you review and adjust them.

1. From the source, open the entity mapping editor to map Raw fields to Curated attributes.
2. Review the AI-suggested mappings and tags, correcting any that are wrong.
3. Save the mapping. Nexa records it as the definition it will compile into pipeline code.

See Curated layer. If two sources describe the same entity, you may also set precedence in Conflict resolution.

Step 5 — Shape the Consumption layer

The Consumption layer (gold) combines curated entities into wide, analysis-ready shapes — for example a 360-degree customer view. You model this visually on the Canvas.

1. Open Canvas and add or edit the consumption entity that joins your curated entities.
2. Map curated attributes into the consumption entity and save.

See Consumption layer and Canvas.

Step 6 — Run the pipeline

Each mapping you saved compiles into a pipeline. Run it to materialize the layers in your data platform.

1. Open Jobs & Pipeline.
2. Find the pipeline generated for your source and run it.
3. Watch the run and check its history to confirm it completes.

See Pipelines overview and Jobs.

What you built

You now have a source flowing through Raw → Curated → Consumption, materialized by a pipeline in your own data platform. From here you can add governance and build on top of the data.

Business glossaryCatalog the business terms behind your entities.

Build an agentPut an AI agent on top of your governed data.

Create an automationAutomate a multi-step workflow with triggers and actions.

Platform architectureSee how the layers fit together conceptually.