A hands-on session on connecting Claude Code to Rankscale and using its REST API and MCP to reproduce the standard Rankscale charts, then build analyses the web interface can't. These pages are the session materials, and a reference to keep afterwards.
What we'll cover
Follow along
Field studies and worked examples we'll refer to during the session. Open them alongside the live demo.
Step-by-step examples from the session - reproducing the standard Rankscale charts through Claude Code, then going beyond with custom metrics you can't build in the web interface.
See the examples → 📊 Field comparisonThe same Rankscale data pulled two ways - the MCP and the REST API - compared on capability, correctness, and token cost. Token cost is what decides which to use at scale.
Read the comparison → 🎯 Variance studyNine AI engines asked the same questions across 1,722 runs: do they return the same brands, in the same order? Where the run-to-run variation comes from.
Read the study →Speakers