zetaris home 2026

Book a demo Start Free

ai_data_gateway / zetaris home 2026

Search

Z

zetaris-cloud

Large node · ap-southeast-2

Provisioning

Provisioning steps

Preparing cloud—
Creating cluster—
Configuring—
Worker nodes—
Components—
Helm charts—
DNS & LB—

Total elapsed: 0s

Your apps & AI agents

Python / Spark SDKJDBC / ODBCMCP / RESTSQL Editor

Zetaris Cloud — your cloud account · your environment · your keys

Query Director Trino · Spark · Presto

Semantic Layer AI Data Steward · context

AgentFlow MCP · orchestration

Data Catalog Iceberg · Delta · metadata

Your data — stays where it is

Snowflake Cloud DWH

Databricks Lakehouse

S3 / Delta Object storage

Postgres Operational DB

On-prem Private cloud

APIs / REST Live feeds

02 · In-situ analytics

One query. Every source stays put.

You've written the same query three different ways this week. One for Snowflake. One for the Postgres instance nobody wants to migrate. One for the S3 bucket that only works if you know the right IAM role. With Zetaris, you define the connection once. Query Director routes each subquery to the right engine at runtime — in-situ, at the source. No pipelines. No copies. No sync lag.

See how Zetaris queries data in place →

Query Director

3 sources · queried in parallel · at the source

Customer Profiles PostgreSQL

Ready Scanning in-situ… ✓ 12,480 rows · 3.1s

Transactions S3 · Delta

Ready Scanning in-situ… ✓ 2.4M rows · 9.6s

Sanctions Watchlist API

Ready Scanning in-situ… ✓ 1,120 rows · 1.2s

Join on customer_id · in-situ

0 records matched

— Success

Claude Connected to Zetaris · MCP

Connecting via Zetaris MCP

Reading unified semantic layer

Generating governed SQL

Running in-situ — no copies

#	Product	Revenue
1	Aurora Wireless Earbuds	$1.28M
2	Nomad Travel Backpack	$1.04M
3	Terra Insulated Bottle	$0.97M
4	Halo Smart Bulb	$0.89M
5	Pulse Fitness Band	$0.83M
6	Drift Desk Lamp	$0.76M
7	Fjord Rain Jacket	$0.71M
8	Ember Coffee Mug	$0.68M
9	Nova Desk Chair	$0.64M
10	Cove Bluetooth Speaker	$0.61M

via Zetaris MCP · live data · no copies

Ask about your data…

Claude by Anthropic

Data for Claude

Point Claude at a Zetaris MCP endpoint. It queries your live data directly — no staging database, no export scripts.

View connection guide →

Codex by OpenAI

Data for Codex

Give Codex a REST endpoint into your actual data. It runs tasks against real schemas — not mocked fixtures or sample exports.

View connection guide →

Cursor AI code editor

Data for Cursor

Add Zetaris as an MCP server in Cursor. Your agent queries live data while you code — explore schemas, run queries, build against real data.

View connection guide →

ChatGPT by OpenAI

Data for ChatGPT

Wire ChatGPT to a Zetaris API endpoint. It gets governed data context on demand — ask questions about your data without exporting anything.

View connection guide →

Trusted by

01 · BYOC

Deploys into your cloud account
IAM role trust model. Zetaris holds zero credentials. Your data never leaves your VPC. Per-customer hardware-backed key vault with automatic rotation.
02 · Self-serve GA

Live in ~20 minutes
Sign up → connect your cloud → launch. One-click Quickstart or guided manual setup. Real-time build log. Full Spark + Trino endpoint ready to accept queries.
03 · Zero data movement

Query across silos. Nothing moves.
Snowflake, Databricks, Postgres, S3, on-prem — one SQL endpoint. In-situ analytics at query time. Define the connection once. Zetaris handles the execution.
04 · AI native

AI Data Steward
Auto-generates data products from your DDL or business glossary. Exposes a governed semantic layer that LLMs and agents query — without touching raw tables.
05 · MCP powered

Unified semantic layer
Unstructured business context — policies, table definitions, domain glossaries — made queryable at runtime. No vector DB required. Every connected agent gets the same governed view.
06 · Open source

Open data catalog + Git
On-the-fly metadata scanning across Delta Lake, Iceberg, CSV, Parquet, Postgres — no ingest step. Full workspace version control with Git. Branch, commit, and roll back your data models.

Free to start Hobby $0.00 Free forever
- Core platform — Spark & Presto engines
- Query Director — Declarative
- Single node · max 12 vCPUs
- No sales call. No procurement. No commitment.
- Live in ~20 minutes · real cluster, your VPC
- You pay your cloud provider for compute. We charge nothing until you upgrade.
Start Free →
Production-ready Professional $0.14 /vCPU/hour
- All Hobby features
- Trino engine + Query Director AI-based
- AI Data Steward · unified semantic layer
- Unlimited CPU scaling · Enterprise Security
- Business hours support · 24hr response
Talk to sales →
Full-scale Enterprise $0.20 /vCPU/hour
- All Professional features
- AgentFlow · Unlimited CPU + GPU scaling
- Enterprise SLAs · dedicated support pod
- 24/7 · Sev1: 4hr · Sev2: 8hr · Sev3: 16hr
- Partner / xSP pricing available
Talk to sales →

The questions your security team will ask.

"We're already on Databricks and Snowflake."

Zetaris sits alongside — it doesn't replace. Connect your Databricks, Snowflake, Postgres, and S3 as sources. Query them together from one Spark/Trino endpoint using the SQL and Iceberg standards you already know. No rip-and-replace, no migration, no retraining.

"We're mid-migration to the cloud."

That's the best time to evaluate us. Zetaris can query both sides simultaneously — legacy on-prem sources and new cloud targets. You remove the deadline pressure of "everything must be migrated before we can query it" and run both in parallel until ready.

"Security won't allow a new cloud vendor into our environment."

Zetaris doesn't live in our cloud — it lives in yours. The cluster runs inside your cloud environment. We use an IAM role trust model with external ID; we never hold a long-lived credential. Your data never leaves your account. The full audit trail stays within your existing compliance boundary.

"Databricks is forcing us to upgrade — we're locked in."

Databricks Standard tier retires October 2026. That's a 35%+ cost increase on every workspace, no opt-out. Zetaris Cloud has no DBU pricing — ever. Flat subscription + vCPU/hour, you pay your cloud provider directly. Spin up a free cluster and run a real workload before you accept that increase.

"My cloud data warehouse can already reach on-prem."

Snowflake and Databricks reach on-prem via S3-compatible storage with constraints. Zetaris gives you three things they can't: talk from a public-cloud Zetaris to a private-cloud Zetaris; use Zetaris as your on-prem-to-cloud data gateway; install in small form factors for edge deployment. The largest deployment choice suite available.