Breaking

Anthropic Details How It Automates 95% of Analytics With Claude

June 3, 2026 at 16:45 EDT

Anthropic has published a new blog post outlining best practices for automating business analytics with Claude, revealing that internally it now automates 95% of its analytics queries via Claude with roughly 95% aggregation accuracy. Announced on June 3 in a post by Claude Developers (@ClaudeDevs), the article "How Anthropic enables self-service data analytics with Claude" explains data-analysis agent building around three pillars: skills, data foundations, and evaluations. Source

June 3 · Anthropic

Skills turn an AI analyst from guesswork into 95% accuracy

Anthropic now automates 95% of its internal analytics queries with Claude — but only after solving what it calls "not a code-generation problem, but a context and verification problem."

The "skills" effect on aggregation accuracy

Same data, same agent — bars drawn to scale.

Without skills

21%

With skills

95%+

Best domains

~99%

95%

of internal analytics queries now automated via Claude

~95%

aggregation accuracy on those automated queries

pillars: skills · data foundations · evaluations

The three-layer stack

1 · Data foundations

Canonical datasets, dimensional modeling, metadata as a first-class citizen.

→

2 · Sources of truth

Semantic layer, lineage and business context.

→

3 · Skills

Procedural knowledge that must call the semantic layer.

Three failure modes the stack targets

Concept ↔ entity ambiguity — a question fails to map to the right field in the data model.
Data staleness — answers built on out-of-date figures.
Retrieval failure — the agent never finds the relevant source.

Continue reading

The rest of this article is for AI News Blitz readers. Choose an option below to keep reading.