Anthropic's Claude automates 95% of internal analytics queries, achieving ~95% accuracy in aggregate.

Anthropic has successfully automated 95% of its internal analytics queries using its AI model, Claude, achieving an impressive ~95% accuracy rate. This innovation allows employees to independently query business data, significantly reducing reliance on data teams.¹²³⁴⁵

Initially, Claude answered only 21% of analytics questions correctly without specific skills. However, after encoding analytical workflows and business context as skills, accuracy improved dramatically, exceeding 95% overall and nearing 99% in certain domains.

The company emphasizes that effective AI analytics hinges on three key principles: maintaining a single source of truth for metrics, ensuring the right data is easily accessible, and continuously identifying stale definitions. This strategic approach not only enhances the efficiency of data handling but also allows the data science team to concentrate on more complex tasks such as causal modeling, forecasting, and machine learning.⁶

As a result, Anthropic is setting a benchmark in the integration of AI into business analytics, showcasing the potential for AI to transform traditional data processes into more streamlined and effective operations.