Skip to content

Analysis

BI ships with it — data does not have to leave to become an answer

Query raw data with SQL, turn the query into a dashboard, publish it behind a link. No exporting to another tool, no waiting in the data team queue.

Real SQL

You write PostgreSQL dialect and the platform translates it to the analytical engine. No proprietary query language.

Publishable artifacts

An analysis becomes a hosted page with a link you can send to a client or an executive.

Refreshes itself

An artifact refreshes when someone looks at it. What nobody opens costs no compute.

Queries

Every source becomes a table in the same query

Checkout sales, ad spend, pixel events, survey answers and your own uploaded datasets are all available as tables. You join them in a single query instead of cross-referencing four exported spreadsheets.

Query result
CampaignSpendRevenueReturn
Retargeting – 7d4.11029.3507,1x
Remarketing – 30d18.24071.9803,9x
Topo / Top of funnel12.87014.2201,1x

Ad spend and checkout revenue on the same row — no spreadsheets exported from two places.

return by campaign
SELECT
  m.campaign_name,
  SUM(m.spend)                        AS spend,
  COUNT(DISTINCT s.transaction_id)    AS sales,
  SUM(s.price_value)                  AS revenue,
  SUM(s.price_value) / SUM(m.spend)   AS roas
FROM meta_ads m
LEFT JOIN checkout_sales s
  ON s.meta_campaign_id = m.campaign_id
WHERE s.status IN ('APPROVED', 'COMPLETE')
GROUP BY 1
ORDER BY revenue DESC

Artifacts

The analysis becomes a page that explains itself

An artifact is a published report: a hosted page with live data behind it. You share the link with a client, a partner or the media team, and it shows the current number — no PDF, no screenshot, no "send me the updated spreadsheet".

  • Public link or restricted to your team
  • A freshness badge showing how current the data is
  • Refresh triggered by visits, not by blind scheduling
  • Version history for every artifact

Dashboards

Dashboards for the routine, queries for the new question

Dashboards cover what you look at every day. When the question nobody planned for shows up — and it always shows up mid-launch — you write the query and get the answer right there, against raw data rather than an aggregate somebody chose to keep six months ago.

Frequently asked questions

Do I need to know SQL?

Not for dashboards. It helps for the new question — and if you do not write SQL, you can ask the AI assistant connected over MCP, which writes the query and returns the result.

Can I connect Power BI or Looker?

Yes, through the REST API and webhook delivery. But the point here is that you should not need them for day-to-day questions.

Will an artifact show stale data?

It always shows how current the data is, and triggers a refresh when someone opens it — so what gets used stays fresh and what got forgotten costs nothing.

Start measuring what actually becomes revenue

Create a free account, connect your first source and see your first lead timeline today. No card required.