Feature
OWOX MCP
Grow your business with data. Get trusted answers you can act on.
In your own AI chat — Claude or ChatGPT. Two prompts, an answer with a chart, every number traceable to SQL your analyst wrote. Zero AI hallucinations.
Get started freeRead the docs →
The analytics system teams trust
Marketplace users
175K+
teams running OWOX from the Google Workspace Marketplace
G2 rating
4.9
296 verified reviews on G2
Connector downloads
156K+
open-source connector installs via npm
GitHub stars
231
the platform is open source — star it, fork it, self-host it
One Feature, Two Perspectives
The analyst governs it. Business teams just ask & get trusted answers.
OWOX Data Marts is one governed surface with two very different experiences — the analyst who sets the boundary, and everyone else who just asks a question within their favorite AI tool.
THE ANALYST
You define what the AI can touch
Bring Data Marts, define the joins, set the quality checks. The LLM only ever reads what you allow – and you see every SQL query it runs, including invented-field attempts, in Run History.
You're the gatekeeper, not the bottleneck.
The Business Teams
You just ask, in your own AI chat
Connect OWOX to Claude or ChatGPT in one click. Ask your business a question in plain English and get an answer within your chat + a connected report in Google Sheets that you can refresh & share.
No dashboard hunting. No waiting on the queue.
Same governed marts. A different experience for every seat. — The analyst keeps control. Everyone else gets the answer, in the tool they already use.
For Data Teams3 use cases
For Data Teams
Govern data marts, not every request
Define the joins and quality checks once. See every AI query — including invented-field attempts — in Run History.
Set guidelines
Publish trusted Data Marts and define the joins — the trust boundary every AI answer respects. Set it up once; every future question inherits the same governance.
Trace SQL for every answer
Every question asked through MCP is logged — including any attempt to invent a field. You stay the gatekeeper, not the bottleneck.
ONLY Structured-SQL
The LLM never writes SQL. It narrates an answer built from your analyst-approved joins — the same question gets the same answer, every time.
For the CEO & AI Lead3 use cases
For the CEO & AI Lead
Ask, and get an answer you trust
Connect in one click, ask in plain English, get a board-ready number with a chart — traceable to SQL your analyst wrote.
Ask in the AI chat you already use
Connect OWOX to Claude or ChatGPT in one click via OAuth. No new tool to learn, no raw warehouse access granted.
Get an answer with a chart
Ask your business a question in plain English and get a board-ready number back — visualized, and traceable to SQL your analyst wrote.
Push it to a Sheet
Keep any answer by sending it straight to a Google Sheet your team already watches — no copy-paste required.
Editions
Included from Senior Analyst and up
Reporting Analyst
Stop being the reporting bottleneck. Hire an Analyst that handles every "quick" question from stakeholders.
Starts at $65 /mo
Hire Reporting AnalystSenior Analyst
Stop running your business blind. Hire a Senior Analyst that unlocks trusted answers to every "why?"
Starts at $90 /mo
Hire Senior Data AnalystEnterprise
Scale self-service analytics with your governance rules and trusted context.
Let's Talk
Talk to SalesWhy teams pick OWOX
Everything business needs
Ask in your own AI chat
No new tool to learn — ask your business a question right inside Claude or ChatGPT, the AI chat you already use.
Answer with a chart, not a guess
Every question comes back with a visualized, board-ready answer — not a wall of numbers to interpret yourself.
Live in weeks, not quarters
No six-month semantic-layer project. Your analyst publishes a mart and defines the keys — MCP is live right after.
Every query logged in Run History
See every question the AI ever asked your data — including any attempt to invent a field. Nothing runs unseen.
An answer in two prompts
Stop waiting on the data queue. Ask now, get a governed answer back in seconds.
Zero hallucinations, by design
The LLM narrates; analyst-approved SQL computes. Every number traces back to SQL your analyst wrote — never invented.
The analyst sets the boundary once
Govern the marts and joins one time. After that, every question in the org runs inside that same trusted boundary.
No dashboard, no ticket, just ask
Skip the backlog. Get the number you need the moment you need it — straight from your own AI chat.
FAQ
Frequently asked questions
Does the AI make up numbers?
No. AI narrates the answer; analyst-approved SQL computes it. The LLM never invents a field or a join.
Is it real-time?
No — MCP answers use the latest scheduled refresh, not a live query. That's deliberate: predictable cost, consistent numbers.
Does data leave our warehouse?
No. OWOX queries your data in place — BigQuery, Snowflake, Databricks, Redshift, or Athena — and never copies it out. The CEO sees analyst-approved governed outputs, not raw data.
Which AI chats work with OWOX MCP?
Claude and ChatGPT today. Cloud-only by design.
Will this replace our analyst?
No. MCP takes the grunt work off their plate. Your analyst keeps the judgment calls — governing the marts and joins — and becomes more valuable, not less.
How long does setup take?
Weeks, not quarters. There's no semantic-layer project first — your analyst publishes a mart and defines the keys.
What users are saying
Every claim has a receipt.
Real things real customers said — each quote pinned to a specific claim, straight from the quotes database.
Nodari RizunFounder & CEO, Pürblack®“AI, by the nature of the models, will hallucinate. And because of that, you need something which will create guardrails to ensure that there are no hallucinations, that you can trust your data.”
Mark SimmonsCMO, Pürblack®“We had six or seven different channels and no single source of truth — it was almost impossible, a lot of guesswork.”
Nodari RizunFounder & CEO, Pürblack®“We regained time. And time is the one resource that never comes back.”