Connects straight to your warehouse. Read-only.

Ask your database a question. Keep the answer.

Dataki writes the SQL, runs it on your warehouse and charts the answer in seconds. Save it as a live dashboard your whole team can open.

Start your free month See it in action

Read-only connections · your credentials never reach the assistant · every query logged.

The AI analytics platform that answers business questions straight from your database.

Connect BigQuery, PostgreSQL, MySQL, Amazon Redshift, Supabase or Google Sheets — or upload a file — and ask in plain English. Dataki writes the SQL, runs it on your warehouse and charts the answer, with the query alongside so anyone can verify it. Save what matters as a live dashboard, or bring Claude, Cursor and ChatGPT in through Dataki's MCP server: they query the same sources without ever seeing a password. Every answer uses your team's definitions, and every dashboard lives where your team can find it.

Reads directly from

  • BigQuery
  • PostgreSQL
  • MySQL
  • Amazon Redshift
  • Supabase
  • Google Sheets
  • CSV and Excel
  • All integrations

01 One question, end to end

Ask in English. Read the SQL. Keep the chart.

Every answer comes with the SQL behind it, visible and editable. Your data team can verify any number in seconds, so everyone else can trust what they read.

Explore the AI assistant
app.dataki.ai / chat bigquery · payments

Which channels brought in revenue last quarter?

Reading the schema, writing the query…
SQL 600 ms · 1.2 GB scanned 
SELECT channel, SUM(amount) AS revenue
FROM   payments
WHERE  created_at >= DATE_SUB(DATE_TRUNC(CURRENT_DATE(), QUARTER), INTERVAL 1 QUARTER)
  AND  created_at <  DATE_TRUNC(CURRENT_DATE(), QUARTER)
GROUP BY channel ORDER BY revenue DESC
Revenue by channel Q3 · 5 rows
Organic searchPartner referralEmailPaid socialDirect $0$100K$200K$300K$400K$500K

02 Dashboards

Every dashboard comes with an analyst built in.

Ask a follow-up right beside the numbers you're reading. Need a new chart, a different breakdown or another filter? Say so, and Dataki adds it to the board.

app.dataki.ai — Revenue overview live · open the chat
Monthly recurring revenue
$164,024
up 7.9% vs last month
MRR by plan
BusinessTeamStarterEnterprise $0$20K$40K$60K$80K

Which plan is actually carrying the growth?

Business. It added $31.4K of the $55.0K gained since June — 57% of net new MRR, from 31% of accounts.

SQL 0.4 s · 4 rows 
SELECT plan,
       SUM(IF(month = '2025-12-01', mrr, -mrr)) AS added
FROM   billing.mrr_by_month
WHERE  month IN ('2025-06-01', '2025-12-01')
GROUP BY plan ORDER BY added DESC
Revenue overview
Filters 01/01/2025 - 31/12/2025
Revenue overview
Monthly recurring revenue
$164,024
up 7.9% vs last month
Active customers
1,284
up 61 vs last month
Net revenue churn
2.1%
down 0.4pp vs last month
Monthly revenue and orders
$0$50K$100K$150K$200K Total Revenue 04008001.2K1.6KTotal Orders JanAprJulOct
MRR by plan
BusinessTeamStarterEnterprise $0$20K$40K$60K$80K
Top accounts
Accounts by MRR
Account Plan MRR Seats Change
Northwind Trading Enterprise $14,200 240 +8.1%
Contoso Labs Business $9,850 118 +12.4%
Fabrikam Health Business $8,400 96 −2.3%
Adventure Works Team $6,120 64 +4.7%
Chat

Which plan is actually carrying the growth?

Comparing plan revenue · 2 steps

Business. It added $31.4K of the $55.0K gained since June — 57% of net new MRR, from 31% of accounts.

SQL 0.4 s · 4 rows 
SELECT plan,
       SUM(IF(month = '2025-12-01', mrr, -mrr)) AS added
FROM   billing.mrr_by_month
WHERE  month IN ('2025-06-01', '2025-12-01')
GROUP BY plan ORDER BY added DESC
Update widget
Type your message…
1

Drag, resize, filter by date, share with a link. Dashboards stay live and up to date, ready for the next meeting.

Explore dashboards

03 Getting your data in

Point it at the database you already have.

Connect in minutes with a read-only login. No new warehouse, no data pipeline, no modelling project before your first answer: Dataki works with the tables you already have.

See every data source
Data sources live · open the menu
A Acme Inc
analytics-prod BigQuery · EU read-only
billing PostgreSQL · 14 tables read-only
warehouse Amazon Redshift · 38 tables read-only
Q4 targets Google Sheets read-only

Dataki connects with a read-only role and encrypts your credentials. The AI model never sees them: it reads your schema, and Dataki runs the query.

04 Security

The model writes the query. It never holds the keys.

Your credentials are encrypted and never enter a prompt. The AI sees your schema, never your passwords. Every query runs read-only and is logged, so you always know who asked what.

See how we keep your data safe
Every answer shows its SQL
Anyone who reads SQL can verify a number in seconds, and edit the query to take it further.
Read-only, always
Dataki connects with a role that can't write. Your data can't be changed, by mistake or otherwise.
Keys you control
Embed and MCP keys come with daily usage limits, and you can revoke any of them instantly.
A full audit trail
Every query is logged with its SQL, its row count and the key that ran it. When someone asks who read what, the answer is ready.
Your company's sign-in
Sign in with Microsoft Entra ID, under your own MFA and access policies.

Most business questions don't need an analyst. They need an answer today. See how Dataki compares.

05 Assistants, over MCP

Ask in Claude. Share it with the whole team.

Your team already asks Claude about the numbers. But every answer lives in one person's chat, built on their own idea of what "revenue" means, and a week later nobody can find it.

Connect Claude, Cursor or ChatGPT to Dataki's MCP server and they query the same sources, with the same definitions, limits and audit log, without ever seeing a credential. In Claude, answers arrive as live Dataki charts built from real query results, and one click adds any of them to a dashboard.

Connect an assistant
Claude Dataki connected

How did August go, and how has revenue moved this year?

Used Dataki: query, render_ui

Dataki's view inside Claude: August revenue of $359K, up 13.65%; 4.29K orders, up 16.07%; an average order of $84, down 2.09%; and a line of monthly revenue rising from about $169K last September to $359K, each with an Add to dashboard button

August closed at $359K, up 13.6% on July on 16% more orders, while the average order slipped 2% to $84. Revenue has more than doubled since last September.

Live in Claude, on Google's public e-commerce sample dataset.

Live dashboards
Saved to Dataki, a chart stays live: it refreshes, takes date ranges and filters, keeps its full history and has a link that lasts.
One definition of revenue
Define your metrics, joins and business terms once. Claude reads them before it writes any SQL, and a correction one person makes can become the rule for everyone.
Organized by team
Sources, definitions and dashboards belong to teams, with owners, admins, members and viewers. Anything Claude saves is shared with the right team automatically, including whoever joins next month.

06 Inside your own product

Put live analytics inside your product.

Embed live dashboards in your app, customer portal or internal tools with two tags. Works with any stack, keeps its own styling and respects your dashboard permissions.

Explore embedding

Embedding a Dataki dashboard took two tags. Our dashboards now live inside our own product, always up to date, and we never had to build a chart ourselves.

Jan Mestek Founder Rented
dashboard.html copy it from any dashboard 
<!-- Load the Dataki embed script once -->
<script type="module" src="https://app.dataki.ai/embed/dataki-embed.es.js"></script>

<!-- Render a dashboard anywhere in your app -->
<dataki-dashboard
  dashboard-id="YOUR_DASHBOARD_ID"
  api-key="YOUR_EMBED_API_KEY"
  theme="dark">
</dataki-dashboard>
Queries and viewers, last 12 weeks your app · dark
08001.6K2.4K3.2K Queries 015304560 Viewers Jul 6Jul 27Aug 17Sep 7

Dataki helps us stay in flow — we can analyze our databases, create dashboards, and share them with the team

Tobias Klimpel CEO & Co-founder MedicalMotion
BigQuery integration
PostgreSQL integration
MySQL integration
Amazon Redshift integration
Supabase integration
Google Sheets integration
+ GA4, Stripe, files…
Via BigQuery exports and uploads
Live Dashboards
Charts & Tables
Embedded Analytics
Metric Alerts

Connect a source → Dataki writes and runs the SQL → a live dashboard

Works with BigQuery, PostgreSQL, MySQL, Amazon Redshift, Supabase, Google Sheets, and CSV and Excel files.

Everyone gets answers. No per-seat pricing.

Everyone on your team can build dashboards, and anyone can view them free. Plans grow with the questions you ask, so rolling Dataki out to the whole company costs nothing extra.

Trial Your first month is free

Sign up with no card and get a month of Dataki on us, then choose the plan that fits. Accounts created before November have full access until 30 November, and everything you build is kept when the trial ends.

Team

€99

per month

First month free

For teams that want answers today, with no analyst queue.

  • Three data sources
  • 100 AI messages a month
  • Unlimited dashboards, unlimited viewers
  • Encrypted credentials no AI agent ever sees
  • Alerts when a metric crosses your threshold
Start your free month

Business

€519

per month

First month free

For the whole company, with the security and control IT expects.

  • Unlimited data sources
  • 600 AI messages a month, then €0.25 each
  • Badge-free embedded dashboards
  • Audit log of every query an agent runs
  • Single sign-on with Microsoft Entra ID
Start your free month

Platform

Custom

priced per app and volume

First month free

Give your customers analytics inside your own product, under your brand.

  • Everything in Business, at Business rates
  • Embed Dataki in one product, for 25 of your customers a month
  • Each customer sees their own data and nobody else's
  • Your branding, themed per customer
  • Dedicated support
Talk to us

An AI message is one question you ask Dataki, however many queries the answer takes. Viewing, refreshing and sharing dashboards, and adding people to your team, are always free.

Team includes 100 messages a month; for more, move up to Business. Business includes 600, then €0.25 per message on your next invoice. Allowances reset every month.

On Platform, a customer counts toward your volume in the months they open a dashboard.

Already on a per-seat or per-editor plan? Your price stays exactly as agreed.

Frequently asked questions

Everything you need to know before you connect your first database.

Can I run queries directly against BigQuery and see the results visualised inline?
Yes. Ask a question in plain English and Dataki writes the SQL, runs it on BigQuery and shows the chart right next to the answer. The SQL behind every chart is one click away, ready to review, edit and re-run. It works the same way with PostgreSQL, MySQL, Amazon Redshift, Supabase and Google Sheets.
Is Dataki a notebook, a BI tool, or something else?
Dataki is AI analytics that combines the speed of a notebook with the reach of a BI tool. You explore by asking one question after another, and every answer is a live chart backed by real SQL. Anything worth keeping becomes a dashboard with a permanent link that refreshes on its own and that anyone in the company can open, no query editor needed.
Can I connect Claude, Cursor or ChatGPT to my own database?
Yes. Dataki runs a remote MCP server, so Claude, Cursor, ChatGPT and any other assistant that supports the Model Context Protocol can list your data sources, run read-only queries and build dashboards. The assistant works with query results, while your database credentials stay encrypted inside Dataki and never reach the model.
Our team already builds dashboards with Claude. Why would we need Dataki?
Dataki turns your assistant's answers into live dashboards the whole company shares. Connected over MCP, Claude saves an answer as a dashboard that re-runs its queries, takes a date range and filters, keeps a history of changes and has its own link. Every source carries one business context for everyone who queries it — metrics as SQL expressions, join paths, columns to avoid and business terms — which Claude, Cursor and ChatGPT read before they write SQL, so everyone works from the same numbers. When someone corrects a definition, they can save it for the whole team. Sources, definitions and dashboards belong to teams with owners, admins, members and viewers, and every dashboard an assistant saves is shared with the team whose data it reads, including people who join later.
Can Claude show Dataki charts inside the conversation?
Yes. In Claude and other hosts that support MCP Apps, Dataki answers with interactive charts, tables and headline numbers, drawn with the same components as a Dataki dashboard. Every number comes straight from the query Dataki ran, and every chart is checked against its data before it appears. One click on Add to dashboard saves any chart as a live widget, on a new dashboard or an existing one. Hosts without MCP Apps get the same answer as text.
Do I need to know SQL to use Dataki?
No. Ask in plain English and Dataki writes and runs the SQL for you. The query is always visible, so anyone who reads SQL can check the logic, adjust it or reuse it anywhere.
Which data sources does Dataki support?
Dataki connects directly to Google BigQuery, PostgreSQL, MySQL, Amazon Redshift, Supabase, Google Sheets and uploaded CSV or Excel files, including hosted databases such as Neon, Amazon RDS, Cloud SQL and PlanetScale. Google Analytics 4, Firebase, Search Console, Firestore and Stripe connect through their BigQuery exports, each with a step-by-step setup guide. Every query runs read-only.
How does Dataki keep AI agents safe around production data?
Every query runs read-only, and every API key acts as the person who created it, so an agent can never reach a source that person could not already open. Keys are scoped by permission, can carry daily limits on query count and bytes scanned, and can be revoked at any time. Every query is logged with the key that ran it, the SQL, the row count and the duration, failures included.
Does Dataki support single sign-on (SSO)?
Yes, with Microsoft Entra ID. Your people sign in with their company account, under your own password policy, MFA and Conditional Access, and land straight in your company's Dataki team. Accounts are matched on IDs your Entra tenant signs, and email addresses are trusted on the domains you list. Setup is two steps: register Dataki in your Entra directory, and we connect it to your team.
Can I embed Dataki dashboards inside my own product?
Yes. The <dataki-dashboard> web component puts any Dataki dashboard inside your app, portal or internal tool with two HTML tags. It works with any framework, keeps its styles separate from your site's CSS, and supports per-customer theming and tenant scoping, so each of your customers sees their own data and nobody else's.
How much does Dataki cost?
Your first month is free from the day you sign up, no card needed. After that, choose a plan: Team at €99 a month (three sources, 100 messages), Business at €519 a month (unlimited sources, 600 messages, then €0.25 each) or Platform, priced per app and volume, for embedding Dataki in your own product. Plans are priced on data sources and AI messages, never per person: everyone on your team can build, and anyone can view a dashboard for free.

Your first month is free

Connect your database. Get your first answer today.

Setup takes about five minutes. Bring a read-only connection and the question your team has been waiting on.

Start your free month Talk to us
Then pick a plan
Team or Business, when your free month ends.
Unlimited dashboards
Shared with as many people as you like.
No credit card
Start today. Nothing to cancel.