Snowflake
×
Papercrane
Dashboards that let your Snowflake warehouse sleep
Papercrane's AI agent writes the SQL, joins your warehouse with GA4, Stripe, or the spreadsheet that never got a pipeline, and publishes hosted dashboards that refresh on your schedule. Between refreshes, viewers read the dashboard, not your warehouse.
Connect in about a minute with your Snowflake credentials. Free to connect and ask. Paid plans start when you publish a dashboard. See pricing
A day in the life of your warehouse
Snowflake bills every second a warehouse is awake, working or idle. How your dashboards are served decides how many of those seconds exist.
Same reporting load, two warehouses
At $3 a credit, that's about $138 versus about $3.
A Medium at 4 credits an hour serving extracts and live viewers, versus scheduled refreshes on an XS. Your sizes and schedules will differ. Ad hoc questions wake a warehouse the same way everywhere; serving viewers is where the days diverge.
You are paying for dashboards three times
The seat bill
The BI contract stacks creator and viewer licenses on top of the warehouse: a mid five figure line before a single credit burns.
The awake bill
Polling, extract refreshes, and the ten minute default suspend timer keep warehouses billing through the idle gaps between queries.
The queue bill
Every question that needs SQL lands on whoever can write it, and half the queue is questions you have already answered.
Ask the questions in your queue
Revenue by customer by month, joining the finance and product schemas in one chart.
Which accounts' usage fell in the 30 days before they churned, and what did they have in common?
Cost per signup by channel: ad spend and GA4 sessions joined to signups in the warehouse.
Stripe fees against invoiced revenue in the ledger schema, monthly, with the gap explained.
The forecast spreadsheet finance keeps outside the warehouse, lined up against actuals inside it.
A usage dashboard for one customer, showing only that customer's rows.
Every answer comes back as a chart on a hosted dashboard, and the SQL is yours to read.
How it works
1. Connect with your Snowflake credentials
2. Grant a role and pick a warehouse
3. The agent plans, writes SQL, and builds a hosted dashboard
4. It refreshes on your schedule, and errors route back to the agent
Papercrane is one more session under the role you grant. Your role hierarchy, row access policies, and masking apply to every query, because Snowflake enforces them, not us.
Give Papercrane its own XS warehouse and a resource monitor if you want a hard ceiling. Multi warehouse support means your dashboards never contend with your loads.
The first dashboard that leaves the account cleanly
Outside your org, Snowflake means reader accounts running on compute you pay for, or exports that go stale in an inbox. A Papercrane dashboard is a hosted page: send a link, embed it in your product, or put it on your own domain, with each customer seeing only their own view.
Sharing is a choice you make per dashboard. Until you publish one, it stays private to your team.
Three ways to put a dashboard on Snowflake
| The job | Snowsight | Seat licensed BI | Snowflake + Papercrane |
|---|---|---|---|
| Stand up a new view | Write the SQL yourself, tile by tile | Model it, then build it, then style it | Ask, review the SQL, publish |
| Who can look at it | Snowflake users holding the driving role | Anyone with a paid seat | People you share it with, no seat required |
| What serving costs | Credits per viewer session | Seats, plus the credits the tool burns | A scheduled refresh; the page serves everyone else |
| A question outside the warehouse | Not its job | Wait for the pipeline to land it | The agent joins GA4, Stripe, or a spreadsheet directly |
| When it breaks | You notice, you fix | A ticket to whoever owns the workbook | Errors route back to the agent, in code you can read |
When you don't need us
If your analysts live in a workbook all day, exploring, pivoting, and passing models between each other, Sigma is built exactly for that motion and is worth its seats. If your questions stay inside Snowflake and you have already invested in Cortex semantic models, Snowflake Intelligence is sitting in your account and speaks your governance natively. Papercrane is worth paying for when the answer needs sources Snowflake does not hold, or viewers who will never have a Snowflake login.
Questions Snowflake teams ask us
The best warehouse questions include data that never landed in the warehouse. Most Snowflake teams join GA4 for traffic, Stripe for payments, and QuickBooks for the close.
Connect your warehouse, grant a role, and ask the question at the top of your queue. The answer comes back as a dashboard your warehouse does not stay awake to serve.
Free to connect and ask. Paid plans start at publishing. See pricing
Papercrane
Integrations: