Free AI readiness assessment

A semantic layer your AI can answer from

Founded by the early Looker team, with 100+ implementations behind us. We build the Omni model, AI context and test questions so answers in Omni AI, Slack or your own agent match the numbers you report.

  • Metrics defined once, so AI and dashboards give the same answer
  • AI context and synonyms written from how your teams talk
  • A test set of real questions, checked before rollout

Request your free AI readiness assessment

With a senior member of our team. You leave with a scope and a timeline. No preparation needed.

Please use your work email so we can prepare for your company.
Enter your name.
Enter your company.
Pick your current BI tool.
Pick a timeline.

A senior member of our team will email you to arrange a time. We use your details to reply to your request. See our privacy policy.

Thanks, we have your request.

A senior member of our team will email you within one business day.

Data teams we have worked with

  • Condé Nast
  • Bloomberg
  • Endeavor
  • Stairwell
  • Zapnito
  • TickPick
  • Laravel
  • Beatstars
  • TRM
  • Thought Industries
  • Central Retail
  • Tread
8,000+hours of Omni delivery
100+Omni implementations
Ex-Lookerfounded by early Looker team members

What the assessment covers

A working session on your model and the questions people ask. Bring whoever owns the metrics.

  1. Your model.The Omni model or dbt project today, and which metrics are defined more than once.
  2. The questions.Real questions from your teams, and which ones AI answers wrongly today.
  3. Scope and timeline.Model work, AI context and testing, as a size band.
  4. Next steps.If it makes sense to go further, we scope the work with you from what we found.

What we build

The layer between your warehouse and every place people ask questions.

Semantic layer

Shared measures, dimensions and topics in Omni, built on your dbt models where you have them.

AI context

Descriptions, synonyms, sample questions and topic instructions, so Omni AI picks the right fields.

Test questions

A set of real questions with agreed answers, run again whenever the model changes.

Access in AI

The same access rules apply to AI answers as to dashboards, tested with real accounts.

How the work runs

Four stages, each with an output you keep.

  1. Collect questions.We gather the questions people ask and the answers the metric owner agrees.
  2. Model.We define each metric once in Omni and remove duplicates.
  3. Context and test.We write the AI context, run the test questions and fix wrong answers in the model or the context.
  4. Roll out.Team by team, with the test set kept as a check for later changes.

Who this is for

  • Omni customers turning on Omni AI for business users
  • Teams whose AI answers do not match the numbers in their dashboards
  • dbt teams that want one set of metric definitions for BI and AI

Who it is not for

  • Anyone looking for an Omni demo, a trial or pricing. Those come from Omni directly.
  • Teams that want a chatbot with no data model behind it

The team behind it

Jeff Garcia

CEO

Spent about eight years at Looker, from 2013 to 2021.

Andrew Searson

Chief Data Officer

Early Looker team member who has led data analytics for 17+ years, building self-service data organizations.

Bruna Baungarten

Director of Delivery

Leads the Shearwater team that runs your project.

Questions

What does the assessment cost, and what do we commit to?

The assessment is free and you commit to nothing. If you go ahead, the migration is scoped as a project from what we learn in the assessment.

Do we need dbt?

No. We can build the model in Omni directly, or on your dbt models if you use dbt, and keep the two in step.

Can the AI show data a user should not see?

AI answers use the same model and access rules as dashboards, and we test them with accounts for different roles.

How do you know the AI is right?

A set of real questions with agreed answers is run before rollout and after each model change. Wrong answers are fixed in the model or the AI context.

Can our own agent use the model?

Yes. Omni exposes the model through its API and MCP server, so your own agent can query the same governed metrics. We can set that up and test it.

Who is Shearwater Data?

An Omni implementation partner founded by the early Looker team. We set up Omni, run migrations to it and support teams once they are live.

Is the assessment a sales call?

It is a working session on your setup with a senior member of our team, which was founded by early Looker team members. If a move makes sense, we offer to scope it. If it does not, we say so.

What do you need from us, and how long is the call?

Nothing in advance. The call runs up to an hour on Google Meet. Bring whoever owns the model and we look at your setup together.

“It’s quite impressive that we were able to swap platforms with Shearwater and still deliver the same experience while cutting costs, all in 4 weeks.”
Jon Beer, COO, Zapnito

See what an AI-ready semantic layer involves

Free, with no commitment. You leave with a scope and a timeline.

Request my free assessment

Shearwater Data is an Omni implementation partner.

© 2026 Shearwater Data. All rights reserved.

Privacy Policy

Thank you! Your submission has been received!
Oops! Something went wrong while submitting the form.