Sculptor is a 📝Clay AI copilot that turns natural-language descriptions into production-ready Clay tables, enrichment workflows, and 📝Claygent prompts — without requiring users to write any prompt syntax by hand.
It exists because Clay's depth — formulas, conditional logic, AI prompts, integrations — historically required hours of practice to wield well; Sculptor collapses that learning curve by letting users describe outcomes instead of writing the syntactic plumbing.
Built for 📝Go-To-Market (GTM), sales-ops, and revenue-ops teams who want Clay's data engine without spending weeks learning its surface area — both new Clay users onboarding for the first time and experienced operators who want to skip boilerplate.
Key Capabilities
- Natural-language prompt building — describe a task in plain conversation and get back a production-ready prompt with variables, output format, and a generated set of test cases.
- Iterative conversational refinement — refine prompts conversationally as you test them on real data, with full version history and instant one-click revert to any prior state.
- Business-context awareness — pulls in your stored Clay context (ICP definition, GTM motion, recent campaign history) so suggestions tune to your strategy instead of a generic template.
- Claygent prompt generation — powers the natural-language build step inside 📝Claygent Builder (Clay), turning agent descriptions into ready-to-deploy Claygent prompts with no manual syntax.
- Table and workflow construction — generates Clay tables, recommends enrichments from the marketplace, and writes formulas from natural-language requests instead of manual configuration.
Getting Started
- Sign in to Clay or sign up at app.clay.com/signup.
- Open any Clay table or workbook — Sculptor is available from the in-app copilot panel.
- Describe in plain language what you want — a new column, a Claygent, a table, an analysis.
- Review Sculptor's generated output and iterate conversationally to refine.
- Approve to apply changes to your table or save the prompt for re-use across the workspace.