Skip to content

Free 30-day pilot

We set it up with you.

The pilot is free for 30 days. It starts with a call and a shared channel with the person who builds Edward, so you never wait in a ticket queue. No data team? We build the catalogue with you.

Rather start alone? The Free plan has the whole product with 200 AI credits a month. During early access we activate new workspaces within 48 hours.

Your first 30 days

A shared Slack or Teams channel

Open for the whole pilot, with us in it. Ask anything, send a screenshot when something looks off, ask for a connector we don't have yet.

  1. Before the call

    A read-only user, made for your database

    We send you a short SQL snippet for your engine. It creates a user that can only read the schemas you pick. No credentials or data are needed before this.

  2. Day 1

    Kickoff call

    On a shared screen, you connect the source and pick the tables. Edward profiles them and drafts names, roles and descriptions. Together we review every field, fix units and aggregations, and hide columns that shouldn't be shown.

  3. Week 1

    Your first dashboards, built with you

    We build the first two or three dashboards with your team, on the questions you ask every week, and check them against numbers you already trust.

  4. Week 2

    Routines and data checks

    We set up Routines so a dashboard page lands in your team's inbox each morning or each week. Warden checks the tables your team approved and emails the data team when one breaks.

  5. Every week

    Check-in

    A short call to see what's used, what's missing and what to add to the catalogue next.

  6. Day 30

    Review

    We look at what the team built and decide together whether to continue.

After day 30

Compare plans →

What we need from you

  • Someone who can create a database user, for about 15 minutes.
  • Someone who knows what the numbers mean, for the catalogue review.
  • Two or three questions your team asks the data every week.

We support PostgreSQL, MySQL, MongoDB, Salesforce, SQL Server, Oracle, Snowflake, BigQuery, Redshift, Iceberg, and CSV or Parquet files.

Under the hood

  • You approve the catalogue

    The AI can only use the tables and columns your team approved. Hidden columns can still join tables but are never shown.

  • The AI picks fields, Edward writes the query

    It asks for catalogue fields, aggregations and filters. Edward writes the query and computes the numbers.

  • Read-only, end to end

    A read-only user on your side, and PostgreSQL, MySQL and MongoDB attached read-only on ours.

How we handle your data →

Plan your pilot

A 30-minute call to scope it on your own data.

Book a demo