Boards on your database, not another SaaS silo

I have watched teams buy a beautiful analytics SaaS, dump CSVs into it, and then spend a year asking for an export. That is not a decision tool. That is a hostage situation.

Atalaya's DaaS stance is the opposite: the board reads your database. If the engagement ends, you still have the tables.

The question, not the wallpaper

A useful board answers one question a person already has. "Are we late on telemetry from line 3?" "Did consent expire for this cohort?" "What did the dealer network do this week?"

Twelve charts that look like a product tour are how you hide that you do not know the question.

We start with a thin slice: one pipeline or one board. You see the shape of the work before a longer engagement.

Wired to the warehouse

The board is a view over models you own. Filters hit fct_ and dim_ tables, not a cached cube we invented so the SaaS feels fast.

That has consequences:

• Freshness is a watermark in the same DB.
• Access control is the same as the rest of your stack.
• GDPR retention is a delete on your rows, not a ticket to a vendor in another region.

GDPR-aware by default

Healthcare and anything with a person in it does not get "user_id" as a joke. Consent, purpose, and retention are columns, not a PDF in a shared drive.

If we cannot say who can see a row and when it must die, we do not put it on a board.

Leave-ability is a feature

You should be able to point Metabase, a custom Next.js app, or a finance export at the same models. That is the test. If the only way to see the number is our UI, we built a silo.

DaaS is the board without you having to hire a platform team. It is not a new source of truth.