AI, SQL and Python. One working notebook.

Ask in plain language, write SQL against your warehouse, or use Python for deeper analysis. Move between chat and code with the queries, results and explanation in one place.

Show me last month's sales breakdown by rep and region, with a trend chart
Called 4 tools ›
Reviewed notebook state
Explored database
Rantitle
Ranrep_region_sales
Called 2 tools ›
Called 2 tools ›
Ask for edits or dive deeper...
Claude Sonnet 4.6
Agentic capabilities built right in
Ask the agent to investigate a question, write the analysis and refine the result with you.

Rep Performance Summary

sql

1SELECT
2 u.id,
3 u.name,
4 u.email,
5 COUNT(o.id) AS order_count,
6 SUM(o.total) AS total_spent
7FROM users u
8LEFT JOIN orders o
9 ON o.user_id = u.id
10WHERE u.created_at >= '2024-01-01'
11 AND u.is_active = true

Order ID	Sales Rep	Region	Deal Stage	Revenue
5001	Sarah Chen	West	Closed Won	$48,500
5002	Marcus Rivera	East	Closed Lost	$22,000
5003	Priya Patel	North	Closed Won	$67,200
5004	Jake Morrison	South	Negotiation	$31,750
5005	Lisa Tanaka	West	Closed Won	$89,000
5006	Derek Owens	East	Closed Won	$45,300
5007	Nina Kowalski	North	Proposal	$53,800
5008	Tom Brennan	South	Closed Won	$71,450

↳rep_summary4

5m ago

Write and inspect SQL

Query your warehouse and inspect the results without leaving the analysis.

Monthly Sales Trend
python
1import altair as alt
2import pandas as pd
3
4monthly_trend_plot_df = monthly_sales_trend.copy()
5monthly_trend_plot_df["sale_month"] =
6pd.to_datetime(monthly_trend_plot_df["sale_month"])
7
8chart_monthly_trend =
9alt.Chart(monthly_trend_plot_df).mark_bar().encode(
Manager performance
↳2trend_chart41m ago
Go further with Python
Transform data, calculate results and build charts in Python.

Guide the agent with what your team knows.

Reference a specific cell, apply a saved skill, or use the definitions and exceptions in your workspace. Give the agent the right starting point for the next question.

Reference cells
Point to the exact result you want to change or explore.
Reuse analytical workflows
Save a repeatable approach so you can use it again.
Bring business rules into the analysis
Use the definitions and exceptions your team has documented.

Rep Performance Summary

sql

1SELECT
2 u.id,
3 u.name,
4 u.email,
5 COUNT(o.id) AS order_count,
6 SUM(o.total) AS total_spent

Order ID	Sales Rep	Region	Deal Stage	Revenue
5001	Sarah Chen	West	Closed Won	$48,500
5002	Marcus Rivera	East	Closed Lost	$22,000
5003	Priya Patel	North	Closed Won	$67,200
5004	Jake Morrison	South	Negotiation	$31,750
5005	Lisa Tanaka	West	Closed Won	$89,000
5006	Derek Owens	East	Closed Won	$45,300
5007	Nina Kowalski	North	Proposal	$53,800
5008	Tom Brennan	South	Closed Won	$71,450

‹12...3442›10 out of 244,241 rows

↳rep_summary4

5m ago

Work in chat or code

Let the agent take the first pass, then open the SQL or Python and make it your own.

Change an input. Update what depends on it.

Queries, Python calculations and charts are connected. Change a filter or an upstream cell and dependent results update together, keeping the analysis in step.

Make the next analysis start smarter.

Capture business rules, trusted queries and repeatable approaches as you work. Keep that knowledge in your shared context so your team and agents can use it again.

Every answer is code you can read

Querio is a glass box. Every AI-generated result is real SQL or Python. You can open the cell, see the join, change the filter, and re-run. No regenerating from scratch, no guessing what the agent did.

Rep Performance Summary

sql

1SELECT

2 u.id,

3 u.name,

4 u.email,

5 COUNT(o.id) AS order_count,

6 SUM(o.total) AS total_spent

7FROM users u

8LEFT JOIN orders o

9 ON o.user_id = u.id

Order ID	Sales Rep	Region	Deal Stage
5001	Sarah Chen	West	Closed Won
5002	Marcus Rivera	East	Closed Lost
5003	Priya Patel	North	Closed Won
5004	Jake Morrison	South	Negotiation
5005	Lisa Tanaka	West	Closed Won
5006	Derek Owens	East	Closed Won

↳rep_summary4

Show me last month's sales breakdown by rep and region, with a trend chart

Called 4 tools ›

Wrote sales_breakdown.md

sql

1SELECT

2 u.id,

3 u.name,

4 u.email,

5 COUNT(o.id) AS order_count,

Ask for edits or dive deeper...

Claude Sonnet 4.6↑

One notebook. More ways to use it.

Arrange notebook outputs into a dashboard, share a report, or schedule the analysis to run again. Keep the underlying code connected to what your team sees.

Turn any notebook into a layout
Create an interactive layout that is always up-to-date.
Sales Last Month+ StatusShare
NotebookLayout
Dunder Mifflin Executive Metrics Board
Core KPIs for Michael's leadership dashboard. All figures use net revenue and exclude cancelled orders.
Month (2026), applies to the breakdowns below
AllJanFebMarAprMayJunJulAug
Monthly Net Revenue Trend
Net Revenue by Paper Type
Net Revenue by Account Manager
Show me our most important metrics at Dunder Mifflin
Worked for 2m ›
I've built an executive metrics board covering KPIs, revenue trends and breakdowns.
Add a filter to view data for different months
month_options = [
"All", "Jan", "Feb",
]
Worked for 1m ›
Updated 4 cells
I've added a month filter at the top of the board, with all of 2026 plus January to August.
Ask for edits or dive deeper...
Build data apps for people who'd rather point and click
Give your team the ability to filter without touching a single line of code.
Share and manage access
Share with your team, make it public for trusted reports, or embed in your product as an iframe.
Automate and run on schedule
Run repeatable analyses on a schedule instead of rebuilding the same report.

Ask a question. Build the analysis. Keep the code.

AI, SQL and Python. One working notebook.

Guide the agent with what your team knows.

Change an input. Update what depends on it.

Make the next analysis start smarter.

Every answer is code you can read

One notebook. More ways to use it.

Build the answer. Keep the reasoning.