Isn't LookML already a semantic layer?

Yes, but one that only Looker can run. We translate it into an open standard (Apache Ossie), so other tools and AI agents use the same definitions.

Yes. Omni uses a comparable code-based model. ParkBee moved from Looker to Omni in 4 weeks with all its dashboards; see Migrate to Omni.

What happens to our derived tables?

We record them with their SQL and dependencies. The quote states whether they move to dbt, the warehouse or the new BI tool.

Looker · LookML

LookML is already code. Now free your logic from Looker.

Your semantics already live in LookML, but only Looker can run them. We translate views, explores and measures into an open standard, so your logic moves with you to any BI tool.

Book a Logic Scan for Looker

Starts with a free 30-minute intake call.

LookML

Years of semantics that only Looker can run

Views and explores

Joins and logic spread over dozens of files

ParkBee

From Looker to Omni in 4 weeks, with 100% of the dashboards

What we extract

What we take out of Looker

Every object that carries logic is read, linked to what depends on it, and checked for use.

Views
Tables and derived fields in LookML, with their SQL.
Explores
How views come together into something users can query.
Measures and dimensions
The definitions behind every number, ready for an open standard.
Derived tables
SQL and native derived tables, and whether they can move to dbt or the warehouse.
Joins
Relationships between views and the conditions behind them.
Looks and dashboards
What users actually open, and what they stopped opening long ago.

What you get

AI-native migration: more than new dashboards

Your logic leaves the old platform as something you own, can read and can query.

Knowledge graph

Every table, procedure, report and relationship in one searchable graph.

Semantic layer in Apache Ossie

Your metrics and dimensions in an open standard, not locked in a tool.

MCP endpoint

Query your logic with Claude, ChatGPT or Copilot.

Plain-language documentation

What every report does and where its numbers come from.

Dead and duplicate objects

What is no longer used does not move.

Reconciliation report

Proof that your core KPIs match the old report.

Optional: a rebuild on Omni, the dbt Semantic Layer, Cube or Power BI.

The result is a semantic layer your BI tools and AI agents share. Choosing one? Read dbt Semantic Layer vs Cube and the Open Semantic Interchange guide. How we expose it to AI: MCP-driven data consulting.

Curious what's in your environment? The Logic Scan shows you in 5 working days.

Book a Logic Scan for Looker

Who does the work

One specialist for Looker

Wesley Nitikromo, 10 years of experience with Looker, LookML and Omni. One fixed point of contact who signs off on the reconciliation.

Proof

What it looks like

“Which reports use finance's revenue definition?”

Apache Ossie

metrics:
  - name: net_revenue
    expression: SUM(amount_excl_vat)
    filters:
      - is_intercompany = false
      - is_credit_note = false
    owner: finance
    source: orders.view.lkml

Case

ParkBee, parking technology

Looker to Omni on Snowflake and dbt · 4 weeks