What is the difference between Root Signals and Scorable?
Root Signals has been rebranded as Scorable. The platform continues to offer the same core functionality for measuring and improving AI-powered applications, including custom AI judges and LLM monitoring, under the new name Scorable.
How does Scorable ensure the reliability of its AI judges?
Scorable calibrates every evaluator against a labeled dataset before production use, providing upfront accuracy metrics. It also minimizes scoring variance compared to raw LLM judges, ensuring consistent and reproducible results for reliable evaluation.
What are the limitations of the free plan?
The free plan includes 1 seat, custom and root evaluators, monitoring, and 100 evaluations per day. Data retention is limited to 6 months, and support is provided via Intercom. It's a good starting point for individual users.
Can Scorable integrate with my existing LLM application?
Yes, Scorable is designed for easy integration. It works with any LLM or framework and can be added to your AI pipeline with a single function call. It supports agents in TypeScript, Python, Proxy, CLI, and cURL.
Does Scorable offer enterprise-level features?
The Scale plan, designed for enterprise, offers unlimited seats, 100,000+ evaluations/month, collaboration features, custom models, on-premise options, SLA support, Slack support, SAML/Okta SSO, and RBAC, with unlimited data retention.