BonnardBonnard

Unlock your Metabase data.

Don't migrate. Just activate. Enable proactive agents to analyse and alert on top of your existing BI data. Powered by an easy-to-use, lightweight semantic layer.

Instant Answers

Stop being the "SQL bottleneck."

Empower non-technical teams to self-serve in team chat. Agents use your curated Topics to answer ad-hoc questions instantly, freeing your analysts for high-value work.

Bonnard

Thread

Sarah K
Sarah KYesterday at 3:56 PM

@Bonnard what organisations do we have?

1 reply
Bonnard
BonnardAPPYesterday at 3:56 PM

Exploring available cards...

what organisations do we have?

Natural Language

Ask "Why is churn up?" and get a visualized answer, not just a link to a dashboard.

Natural Language
SK

Why is churn up this month?

B

Churn increased 18% this month, driven primarily by the Enterprise segment...

Ask questions in plain English

High Accuracy

Answers are grounded strictly in your curated data, showing the SQL generated for full transparency.

Generated Query
✓ Verified
SELECT
segment, COUNT(*) as churned
FROM
customers
WHERE
status = 'churned'
GROUP BY
segment
Every answer shows the SQL for transparency

Drill Down

Users can ask follow-up questions to slice and dice data without needing to learn the BI tool.

Follow-up Thread
→Break down by region?

EMEA: 42%, APAC: 31%, Americas: 27%

→Why is EMEA so high?
Analyzing...
Keep asking to dig deeper

Topics

A Lightweight Semantic Layer.

Data sprawl happens. Manage it effectively with Topics. Designate specific data clusters as trusted sources for AI and define business rules using natural language. Achieve clarity and governance without engineering a new semantic layer from scratch.

Bonnard
Revenue & Finance
Description
Financial metrics and revenue analysis

A brief description to help users understand what data this topic contains

Business Context
MRR: Sum of all active monthly subscriptions
Churn: Cancelled or not renewed within 30 days
NRR: Includes expansion from upsells/cross-sells
Segments: Enterprise ($50k+), Mid-Market ($10-50k), SMB (<$10k)

Detailed context helps AI agents understand how to use and interpret the data in this topic.

Cards (7)
Revenue by Segment
Monthly MRR Movement
Customer Acquisition Cost
Active Subscriptions
Churn by Cohort
Pipeline Coverage
Net Revenue Retention
User Access (1)
Sarah K@sarah

Curated Universes

Prevent the AI from getting lost by restricting it to specific, high-quality datasets for each team.

Cards (7)
Revenue by Segment
MRR Movement
Customer Cohorts
Churn Analysis
Restrict AI to high-quality, curated data

Business Context

Define terms like "Active User" or "Churn" in plain English once, and the agent respects it forever.

Business Context
"Active User" means a user with at least one session in the last 30 days. "Churn" refers to subscriptions cancelled or expired without renewal. MRR is calculated as the sum of all active monthly subscription amounts.
Detailed context helps AI understand how to use and interpret the data in this topic.
User Access (2)
Sarah K@sarah

Deep Analysis Reports

Go beyond the "what." Understand the "why."

Most tools just run SQL and email a chart. Bonnard agents perform cross-question analysis—synthesizing data from multiple dashboards to explain drivers, trends, and causality automatically.

Bonnard
Weekly Revenue Analysis
Sent every Monday at 9am
Data Source Type
Include dashboards or individual questions
Search questions...
Selected Questions (4)
Revenue by Segment
Monthly breakdown of revenue across customer segments
Model • postgres • ID: 42
Customer Acquisition Cost
CAC trends with channel attribution
Model • postgres • ID: 87
Churn Analysis
Churned accounts by reason and cohort
Model • postgres • ID: 103
Pipeline Coverage
Current quarter pipeline vs target
Model • postgres • ID: 156

Institutional Memory

Bonnard reads your existing metadata, descriptions, and relationships. It learns from what you've already built.

Connected Sources
Synced
orders
12 columns • 3 joins detected
customers
8 columns • 2 joins detected
subscriptions
6 columns • 1 join detected
Bonnard reads your Metabase metadata automatically

Micro-Analyses

Instruct agents to perform multiple layers of investigation within a single report to build a complete narrative.

Analysis Steps
1
Summarize revenue by segment
Query: Revenue by Segment
2
Identify top growth drivers
Compare to previous period
3
Explain any anomalies
Flag significant changes
+
Add another analysis step

Meeting-Ready

Reports are delivered in rich formats to Slack, Teams, or Email, ready to be forwarded to leadership immediately.

Delivery Channels
3 active
#leadership-updates
Slack • Every Monday 9am
exec-team@company.com
Email • Every Monday 9am
Finance Team
Microsoft Teams • On demand
Rich formatting, ready to forward to stakeholders

Try Bonnard now.