Authenticate
All services use bearer tokens. See Authentication for
the token types and a working curl example.
API Overview
Nexa exposes its platform through a set of REST services. Use them to automate everything the
web UI (nexa-web) does — catalog operations, code generation, pipeline and job orchestration,
lineage, agents, and agentic automations — from your own scripts, CI jobs, or the Nexa CLI. This
page maps the services so you know which one to call; each has a full, auto-generated OpenAPI
reference linked below.
Because Nexa is customer-managed (DataReadyAI ships the container images and IaC; your platform team runs them in your own AWS or Azure account against Snowflake or Databricks), there is no single public base URL. The host is whatever your deployment exposes for each service. What is stable is the path prefix each service serves under, listed below.
The services
Section titled “The services”The platform is a set of independently deployed services. The web app talks to nexa-backend
(TypeScript), which fronts three FastAPI data/AI services. You typically call nexa-backend from
the UI, but the FastAPI services are directly addressable and are what the generated references
document.
| Service | Repo | What it does | Path prefix | Reference |
|---|---|---|---|---|
| Databricks Platform API | nexa-databricks-api |
Catalog, code generation, pipelines, jobs, lineage, SQL execution, deployments on Databricks | /dbx + /api/v1/… |
Databricks API |
| Snowflake API | nexa-snowflake-api |
The same surface for Snowflake — catalog, codegen, connectors, pipelines | /api/v1/… |
Snowflake API |
| Agents API | nexa-agents-api |
Author, validate, deploy, and run agents and automations; Slack bindings | /api/agents/… |
Agents API |
| ACTG-AI | actg-ai |
Async LLM job engine behind mapping, conflict resolution, and codegen | /api/v1/… |
ACTG-AI Service |
The Databricks and Snowflake services are two implementations of the same conceptual API — one per data platform. Which one is live depends on how your environment was installed. Prefer synced platform tabs in your tooling rather than assuming one.
How the services relate
Section titled “How the services relate”nexa-backendis the orchestrator the UI calls. It holds environment configuration (which data platform, which Git provider) and forwards work to the right data-platform API.- The Databricks / Snowflake API owns Unity Catalog / Snowflake catalog metadata, generates DDL and ETL, and runs pipelines and jobs. Long-running LLM work (mapping, conflict resolution, projections) is delegated to ACTG-AI as asynchronous jobs.
- The Agents API is a separate lifecycle service for building and running agents and automations. It reads the same environments and Git integration but manages its own agent registry and deployment state.
Reference documentation
Section titled “Reference documentation”Each reference below is generated directly from the service’s OpenAPI spec, so it always matches the deployed version. Start there for exact request/response schemas and the full endpoint list.
Databricks Platform APICatalog, codegen, pipelines, jobs, lineage, SQL, deployments.
Snowflake APIThe Snowflake implementation of the data-platform surface.
Agents APIAgent and automation authoring, deployment, and runs.
ACTG-AI ServiceThe async LLM job engine (full reference generated on request).
Next steps
Section titled “Next steps”Learn the conventions
Shared concepts — environments, errors, request IDs, and versioning — apply across every service.
Use the CLI
The Nexa CLI wraps these APIs with one authenticated, environment-aware command surface.