Skip to content

Introduction

Nexa is a data and AI platform by DataReadyAI that you run inside your own cloud account. This section is for the platform team that installs, operates, and secures Nexa — not the analysts and data engineers who use it day to day. Start here to understand what gets deployed, what it depends on, and what to prepare before your first install.

Nexa ingests raw data, maps it through Raw → Curated → Consumption layers (bronze/silver/gold), generates the pipelines that move data between those layers, and runs agents and agentic automations on top — with governance built in (business glossary, PII tagging, change review, cost tracking). It does not store your data itself: it orchestrates your existing data warehouse (Snowflake or Databricks), where the tables actually live.

For the platform team, the measurable outcome is a single, governed control plane for the whole data lifecycle — source onboarding, pipeline generation, and AI agents — deployed as a set of container images and Infrastructure-as-Code (IaC) you own and can audit, with no data leaving your environment for DataReadyAI to see.

Nexa is a control and orchestration layer. It sits between your data sources and your warehouse, and it drives the warehouse rather than replacing it.

Data sources Nexa (in your cloud) Your warehouse
┌──────────────┐ ┌─────────────────────────┐ ┌──────────────────┐
│ Postgres │ │ nexa-web (UI) │ │ Databricks │
│ Kafka │ ──────► │ nexa-backend (API) │ ─────► │ or │
│ S3 / volumes │ │ warehouse APIs │ │ Snowflake │
│ files, apps │ │ agents-api + workers │ │ (Raw/Curated/ │
└──────────────┘ └─────────────────────────┘ │ Consumption) │
runs on your EKS/AKS cluster └──────────────────┘

Nexa augments your warehouse and Git/CI rather than replacing them: it generates pipeline code and pushes it through your GitHub repositories, and it executes work as jobs and pipelines in Snowflake or Databricks. It replaces ad-hoc, hand-written ingestion and modeling glue, and the spreadsheets teams otherwise use to track glossary, lineage, and cost.

Nexa is customer-managed. DataReadyAI ships container images (to your registry) plus Terraform/Terragrunt IaC and Kubernetes manifests; your platform team installs and operates it. DataReadyAI provides support but has no standing access to your environment. See Deployment model for the responsibility split.

At a high level you will:

  1. Provision cloud infrastructure with the supplied IaC — a Kubernetes cluster (Amazon EKS or Azure AKS), a container registry, DNS, a managed Postgres for Nexa’s own metadata, and networking.
  2. Deploy the Nexa services to the cluster via GitOps (ArgoCD + Kustomize).
  3. Connect Nexa to your Snowflake or Databricks workspace and to GitHub.
Path Who it’s for Where to go
Evaluate the product Trying Nexa’s features before committing infra Platform overview and Quickstart
Full install Platform team standing up a real environment PrerequisitesInstallation
Understand the footprint first Security/architecture reviewers Reference architecture
AWS Deploying on Amazon EKS Prerequisites — AWS tab
Azure Deploying on Azure AKS Prerequisites — Azure tab
Snowflake Warehouse is Snowflake Prerequisites — Snowflake tab
Databricks Warehouse is Databricks Prerequisites — Databricks tab