rigami.
All stacks
/// AI · Frontier LLM

Claude Sonnet 4.5 for tool-using agents.

Claude is our default for agentic, long-context, tool-using workflows. We build production Claude apps with the eval and observability layer wrapped around them.

Start a Anthropic Claude project AI & LLM Products
/// What we ship with Anthropic Claude

How we use it
in production.

  • Claude Sonnet 4.5 / Haiku 4.5 / Opus 4.5
  • Tool-calling and computer use
  • Long-context reasoning over docs
  • Agentic workflows with retries and budgets
  • Prompt caching for cost control
  • PII redaction and safety layering
/// Anthropic Claude in production

Where it shipped.

Wholosophy

AI-powered wellness coaching that meets people where they are.

Case study
RealProfits

A retail trading and profits-tracking platform with an AI co-pilot.

Case study