DATARIG LITE BY EXASCALE — FREE FOREVER

The AI data engineer that runs on your machine.

DataRig Lite is free for your desktop: describe a pipeline in plain English, review exactly what it builds, then approve — governed Iceberg tables, orchestrated dbt models, and live dashboards, without your data ever leaving your laptop.

Download free for Windows See how it works

Zero data retention — nothing leaves your machine
Review-gated agents — nothing runs without approval
Open Iceberg tables — no lock-in
All tools built into the rig

Part of the ExaScale DataRig platform: start free on your desktop, scale to managed Cloud, or deploy in your own VPC — same pipelines, same governance, zero rewrites.

DataRig — localhost:8088 local runtime healthy

You Load orders.csv into Iceberg, then build a daily revenue dashboard.

PROPOSAL · REVIEW REQUIRED sha256 : 9f41c…e2b7

Source analysis

orders.csvCSV · 4,213 rows12 columnstypes inferred ✓

Plan

01 Spark ingestion → local.user.orders (Iceberg · Polaris)
02 dbt model revenue_daily + freshness & row-count tests
03 Airflow DAG orders_daily, scheduled 06:00
04 Superset dashboard “Daily Revenue” over Trino

✓ Completed Iceberg table created dbt tests passed Dashboard live

Everything you need, packaged in one install. Nothing hand-wired.

Apache Iceberg
Apache Spark
Trino
Apache Polaris
Kafka + Debezium
Apache Airflow
dbt Core
Apache Superset

Why DataRig

A real pipeline stack, without the platform tax

Proprietary platforms charge you rent to write pipelines yourself. Hand-rolled open source costs you a weekend a month. DataRig removes both problems.

⌂

Local-first, zero data retention

Everything runs on your machine. Datasets, prompts, schemas, and logs are never sent to ExaScale. Local AI via Ollama by default; remote telemetry is off by default.

✓

Review-gated agents, not autopilot

Every generated artifact — ingestion specs, DAGs, dbt models, SQL, dashboards — is validated and hash-bound. Nothing executes until you approve the exact plan.

◈

Open formats, zero lock-in

Your tables are Apache Iceberg v2 behind an open REST catalog. Leave anytime and your data comes with you in standard formats any engine can read.

◎

Every tool built into the rig

SQL explorer, orchestration, dbt docs, dashboards, CDC connectors — pre-connected and one click away. You never assemble seven projects by hand.

✦

AI that knows your data

DataRig profiles your sources, infers types, and drafts plans grounded in registered adapters — a tool name in a prompt is never mistaken for executability.

Free while it proves itself

DataRig Lite is free forever — complete, and local-first, not a crippled trial. Paid managed editions come later, for teams that outgrow their laptop.

How it works

From plain English to production, in four steps

You stay in control: DataRig proposes the exact plan, you approve the exact hash, your machine runs it.

STEP 01
Connect

Drop in CSV, JSON, or Parquet files — or connect PostgreSQL / SQL Server for continuous CDC. All local.
STEP 02
Describe

Say what you want: “load this into Iceberg,” “add dbt tests,” “build a revenue dashboard.”
STEP 03
Review & approve

Inspect source analysis, plan steps, and generated files. Approve binds to an exact content hash.
STEP 04
Run & monitor

Watch live Spark logs, staged progress, and status chips — with error-specific recovery if anything fails.

AI Data Explorer

Explore governed data, without leaving the rig

Browse your Iceberg catalog tree, inspect schemas, and run read-only SQL — right beside the chat. AI-suggested queries load here for review; they never auto-run.

ON THE ROADMAP Curated metadata → accurate NL-to-SQL

Annotate your tables and columns once; AI-drafted SQL gets dramatically more accurate. Every draft still passes a validation layer — statement allowlists, reference resolution, forced row limits — before it can run.

ON THE ROADMAP Template chips & semantic layer

Deterministic, zero-model starting points for common queries, built on your governed table metadata — the reliable path to answers before any AI is involved.

CATALOG

▾ local ▾ user · orders▸ customers· order_items· revenue_daily

SQL · READ-ONLY · 200 ROWS MAX

SELECT
  order_date,
  SUM(amount) AS revenue
FROM local.user.orders
GROUP BY order_date
ORDER BY order_date DESC;

SCHEMA

order_id BIGINT customer_id BIGINT order_date DATE amount DECIMAL(12,2)

RESULT · 7 ROWS

order_date	revenue
2026-08-25	18,204.50
2026-08-24	16,930.00
2026-08-23	21,417.75
2026-08-22	14,882.10

Preview capped at 200 rows · full power via Trino or DBeaver

Built into the rig

All the tools, pre-connected

Specialist consoles are packaged, wired, and one click away — with the DataRig console as your single home base.

Trino SQL engine

Airflow Orchestration

dbt Docs Models & lineage

Superset Dashboards

Kafka Connect CDC pipeline

Grafana Runtime health

Capabilities

The whole pipeline, one conversation

Sources → Iceberg → Airflow → dbt → Superset. Each stage is a governed skill with its own validated contract.

⇥

Ingestion

Land CSV, JSON, JSONL, and Parquet files — or HTTPS APIs — into governed Iceberg tables. Full load today; XML profiling is review-ready.

⇄

CDC & streaming

Capture changes from PostgreSQL and SQL Server with Kafka + Debezium. Snapshot, updates, deletes, restart recovery — verified end to end.

⚙

Transformation

dbt models generated with tests and docs, executed by Airflow against Trino — every artifact reviewable before it runs.

◷

Orchestration

Airflow DAGs created, scheduled, paused, retried, and refreshed in place. Run-scoped docs and live status inside the console.

✓

Data quality

Freshness checks, row-count parity, and null/duplicate tests proposed as part of the plan — not bolted on afterwards.

◎

Analytics & BI

Superset datasets, charts, and dashboards provisioned from your governed business layer, queried through Trino.

One platform · three runtimes

Start on your laptop. Never start over.

The pipelines you build in Lite are the same artifacts every edition runs: Iceberg tables, Airflow DAGs, tested dbt models, Superset dashboards. When you scale, your work scales with you — no rewrites, no migration project.

LITE · FREE FOREVER

Your machine

Individual engineers, pilots, local-first teams

Complete local lakehouse + review-gated AI
All tools built in — zero setup beyond install
Zero data retention, telemetry off

$0 forever

Download Lite

CLOUD · COMING SOON

Our cloud

Teams that outgrew the laptop but not their budget

Everything in Lite — fully managed and always on
Automated upgrades, backups, and monitoring
Transparent usage metering — no opaque credits, no idle cluster sprawl
Email + Slack support

Usage-based pay for what you use

Join the waitlist

ENTERPRISE · COMING SOON

Your VPC

Regulated orgs that need sovereign control

Everything in Cloud, deployed inside your network perimeter
SSO, organization RBAC, audit export
Customer-managed encryption + private networking
Contractual SLA with dedicated support

Annual contract + SLA

Pipeline definitions carry forward across every edition. Your data stays in open Apache Iceberg formats everywhere.

On the roadmap · multi-tool connectivity

Every platform locks its AI inside its own wall

Cortex lives in SnowflakeGenie lives in DatabricksCopilot lives in FabricGemini lives in BigQuery

ExaScale is the bridge when your data
doesn't live in one wall.

⇄

Query across warehouses

Planned: read-only federation to Snowflake, Redshift, Databricks, and BigQuery catalogs — join lakehouse and warehouse data without ETL sprawl, behind the same review-gate stack.

◈

Bring your semantics

Planned: import existing semantic models and metadata from the platforms you already pay for — your definitions travel with you instead of being rebuilt per vendor.

✓

Gated write-back, last

Federation arrives read-only first. Any future write-back adapter lands behind explicit, hash-bound approvals — never silent synchronization.

Designed, not yet scheduled — join the waitlist to shape the priority.

Built for people who own data

Whatever you run today, DataRig takes the toil

Data & analytics engineers

Stop hand-wiring load scripts and cron jobs. Describe the pipeline, review the plan, ship the dashboard — all locally.

Analytics leads & BI owners

Let your team self-serve pipelines safely. Review gates mean nothing mutates data without an approved plan.

Technical founders & CTOs

A governed lakehouse on a laptop, with zero data egress and open Iceberg formats — no vendor bill, no lock-in.

Analytics leads on a mixed stack

Your team pays for AI assistants in two platforms and still can’t join the data. DataRig is planned to bridge warehouses and lakehouse — read-only first, always gated.

Editions

Start free with Lite. Scale to the cloud when you're ready.

DataRig Lite is the complete local product — free forever. Paid editions exist for teams that want managed operations, not extra features on your desktop.

DATARIG LITE · FREE FOREVER

Lite — for your machine

Complete pipeline: files/CDC → Iceberg → Airflow → dbt → Superset
AI Data Explorer + review-gated agents
All tools built in, pre-connected, localhost-only
Local AI (Ollama) by default · zero data retention · telemetry off
Windows & macOS, no cloud account needed

Download Lite free

MANAGED · COMING SOON

Cloud — we run it

Usage/based

Everything in Lite — fully managed by ExaScale in dedicated cloud projects
Automated upgrades, backups & monitoring
Transparent usage metering — no DBU-style opaque credits
No idle cluster sprawl quietly burning your bill
Frontier model providers when you choose them
Email + Slack support