Comparison
Codd AI vs AtScale
AtScale is a universal semantic layer with deep roots in OLAP and BI, strong at serving governed metrics to Excel, Power BI and Tableau. Codd AI is an AI-native contextual semantic layer: it uses AI to discover your business knowledge and generate the data model and metrics, keeps people in the loop to certify them, and delivers answers through conversational analytics, AI agents and your BI tools. Choose AtScale for hand-modeled, BI-centric semantic models at scale; choose Codd AI when the bottleneck is capturing business context fast enough for AI.
Last reviewed October 2026, based on public documentation. Spotted something out of date? Let us know.
At a glance
| Capability | Codd AI | AtScale |
|---|---|---|
| Core approach | AI-native contextual semantic layer: business knowledge, rules and metrics in one governed layer | Universal semantic layer with virtual OLAP heritage and an AI computation engine |
| How the model is built | AI discovers entities, relationships and metric logic from data, BI assets, documents and experts; people review and certify | Modeled in SML (open, YAML-based, versioned in Git) or the visual Design Center; AI-assisted modeling (CoModel) announced |
| Business knowledge beyond the database | Built in: captures definitions and rules from documents, BI reports and subject-matter experts | Focused on the semantic model; knowledge outside the model is maintained separately |
| BI tools | Power BI, Tableau, Looker, Excel | Excel, Power BI, Tableau, Looker and others via MDX, DAX, SQL and XMLA |
| AI agents and LLMs | Conversational analytics built in, plus MCP, Slack and Teams; context works with any LLM | MCP server that gives agents governed metrics; the LLM or agent writes the questions |
| Repeatable analysis and automation | Playbooks for repeatable, auditable AI reasoning and Actions for scheduled, human-approved execution | Not a focus; metrics are served to the tools and agents that consume them |
| Data platforms | Snowflake, Databricks, BigQuery, Redshift, PostgreSQL, Oracle, Teradata, Trino | Snowflake, Databricks, BigQuery, Redshift, Azure, Apache Iceberg |
| Query performance | Queries run in your warehouse | Queries run in your warehouse, with automatically managed aggregates |
| Pricing | Contact sales | Consumption-based, by quote |
What is AtScale?
AtScale describes itself as a universal semantic layer. It sits on top of your cloud data platform and lets BI tools, spreadsheets and AI agents query one governed set of business metrics without moving data.
Models are defined in SML, an open YAML-based modeling language versioned in Git, or in a visual designer. AtScale translates MDX, DAX and SQL queries into queries against your warehouse and can manage aggregate tables there to speed up common queries.
For AI, AtScale offers an MCP server so agents can request certified metrics by name, and it has announced AI-assisted modeling that routes proposed changes to data engineers for approval.
Key differences
Who builds the semantic layer, and how fast
With AtScale, data engineers model the business in SML or the visual designer, which suits teams that want full manual control. AtScale cites a first governed model in roughly six to eight weeks. Codd AI starts by discovering what your business already knows, across data, existing BI assets, documents and experts, and generates the ontology, data model and metrics for people to review and certify. The goal is to remove the slowest step: getting business context out of people's heads and into a governed system.
Metrics vs business context
A metric definition tells a tool how to calculate revenue. Business context also tells an AI why a metric exists, which exceptions apply and how terms differ between departments. AtScale focuses on governed metrics for BI and agents. Codd AI is built as a contextual semantic layer so AI can reason about the business, not only query it. See what is contextual analytics.
BI-first vs AI-first
AtScale's OLAP heritage makes it a strong fit for large Excel and Power BI estates that depend on MDX and DAX. Codd AI is designed AI-first: conversational analytics, agents, Playbooks and Actions sit on the same certified context, while BI tools consume the same definitions.
Which should you choose?
Choose AtScale if
- Your priority is serving governed metrics to a large Excel, Power BI or Tableau estate, including MDX and DAX clients.
- Your data engineering team wants to hand-model everything in code (SML) with a Git workflow.
- Query acceleration through managed aggregates is a key requirement.
Choose Codd AI if
- The bottleneck is capturing business definitions and rules, not writing models.
- You want AI to generate the semantic layer, with people certifying it, in weeks rather than quarters.
- You need conversational analytics, AI agents, repeatable Playbooks and Actions on one trusted context.
- Important knowledge lives in documents, BI reports and experts, not only in the database.
Frequently asked questions
What is the main difference between Codd AI and AtScale?
AtScale is a universal semantic layer that teams model by hand in SML or a visual designer, strong at serving governed metrics to BI tools. Codd AI uses AI to discover business knowledge and generate the semantic layer, with human certification, and is built for conversational analytics and AI agents.
Does AtScale support AI agents?
Yes. AtScale provides an MCP server that lets AI agents request governed metrics from its semantic models. Codd AI also supports MCP and adds built-in conversational analytics, Playbooks and Actions on the same certified context.
Can Codd AI replace AtScale?
For teams whose priority is AI analytics and business context, yes. Teams that rely heavily on MDX and DAX clients or managed aggregates may prefer AtScale for those workloads. The right choice depends on whether your main gap is BI acceleration or business context for AI.
See Codd AI on your own data
Learn how Codd AI builds a certified, contextual semantic layer in weeks, and how it fits alongside the tools you already use.