- ObserveAgents use your data
- DiagnoseBonnard finds what’s wrong
- ImproveYour data team answers, into the context
- VerifyEvals check it helped
Bring the context you have
Connect Postgres or your warehouse, and Bonnard reads the tables, keys and joins you already have. Add a dbt or OSI file if you have one.
- Start hereYour database or warehouseTables, keys, comments. Connected read-only.
- OptionalA dbt or OSI fileDescriptions, keys, joins, metrics.
- WinsYour team, through your agentDefinitions, metrics, rules. Your edits win.
public.ordersWhat your agent reads- DescriptionOne row per checkout. Test orders are excluded.Your team
Replaced: Orders placed through checkout.dbt or OSI file - KeyidDatabase
- Time columnplaced_atYour team
- Joincustomers on customer_idDatabase
- Joinrefunds on order_iddbt or OSI file
Your agent does the work
Define metrics, rules, evals and metadata from the agent that already knows your business. Draft it through the MCP, review it, then publish it.
Add to the orders description that cancelled orders are included.
edit_draftIn the draft, nobody is served itLooks right, publish it.
publish_draftPublished version 3An admin decides whether AI clients can publish. Roll back in the console.
What it can edit
Metrics
“Net revenue is paid orders, minus refunds.”
Glossary
“A customer is a firm, a client is a person.”
Rules for every query
“No billing data here. Say it’s unavailable, don’t estimate it.”
Descriptions
“One row per checkout. Test orders are excluded.”
Keys and joins
“orders joins customers on customer_id.”
Searchable values
“org_name holds our customers’ names.”
Hidden tables
“Hide orders_old. It stopped updating in May.”
Eval questions
“Add net revenue last month to the golden set.”
Evals built inComing soon
Run your questions against the live context and any draft, and see what got better, question by question. Then keep running them, so you know your agents are still right.
- 01
Pick the questions
From your verified dashboards, or generated per team.
- 02
Run any version
About 5 answers per question, on live, a draft or a past version.
- 03
Compare
Question by question, scored against your verified dashboard.
- 04
Keep running
Weekly and after each publish, so you see accuracy slip.
| Question | Live | Draft | Result | Why it changed |
|---|---|---|---|---|
| What was net revenue last month? | Better | Net revenue now has one definition | ||
| What is our average order value? | Better | Uses the revenue metric | ||
| What share of orders were refunded? | No clear difference | Within the run-to-run spread |
The context loop
Bonnard reads how agents used your data, finds what’s wrong and asks your data team. Their answers go into the context.
- 01
Observe
The activity shows 8 people using 3 definitions.
Bonnard activityLast 30 daysInconsistent definitionsnet revenue, 3 definitions, 8 peopleQuestionsum(total)[status = paid]5 peoplesum(total - refunds)[status = paid]2 peoplesum(total)1 person
- 02
Diagnose
Your agent asks what the data can’t settle.
People use three definitions of net revenue. The data can’t settle it. Which one is right?
Paid orders, minus refunds.
- 03
Improve
The answer goes into a draft, and a person publishes it.
okedit_draftNet revenue added to the draftAlex MorganPublished version 4LiveMetricnet_revenuePaid orders, minus refunds. - 04
Verify
Coming soonEvals check it helped.
What was net revenue last month?BetterNet revenue now has one definition
It also finds
- Searches that found nothing
- Struggling work
- Errors
- Stale or misleading data
Start scaling
accurate data today
Free for your first 3,000 calls. No seat limits.
