Introducing Decagon Labs.

Testing & experimentation

No items found.

No items found.

Why we built Decagon Duet on Agent Operating Procedures

Optimizing GEPA for production: A test-driven approach to prompt engineering

Bayesian VAD: Fixing turn detection in production voice pipelines

Introducing Decagon Labs

Introducing Templates: Launch faster with proven use cases for AOPs, tools, and Watchtower

The QA breakthrough: Watchtower turns every support interaction into actionable intelligence

Climb higher: Introducing Diagnostic Tools for faster agent iteration

Trace View: AI agents shouldn’t be black boxes

Engineering fast, performant AI agents through fine-tuning

Introducing Agent Versioning: engineering-grade governance for AI agents

Simulations: How AI agents get tested and trusted

The future of AI agents is test-driven

The evaluation engine behind Decagon’s AI agents

Introducing Decagon’s experimentation & A/B testing suite

⁨Explore more topics

AI agent building

Testing & experimentation

Analytics & VoC

Voice & omnichannel support

Guardrails, security, & governance

Use cases & experiences

Workplace

The future of customer experience starts here.