AI agents face the ROI test: Sierra co-founder Clay Bavor

Agentic AI represents a shift toward software that can reason, make decisions, and execute complex tasks autonomously, particularly within customer service environments. Unlike traditional subscription-based models, outcome-based pricing aligns developer incentives with business success by charging only when specific goals, such as resolving a customer inquiry or retaining a client, are achieved. While frontier models provide the core intelligence, specialized platforms like Sierra orchestrate a "constellation" of models to optimize for performance, latency, and cost-effectiveness in regulated industries. The current enterprise focus on return on investment has moved beyond simple token usage toward evaluating tangible business outcomes. As coding agents demonstrate significant productivity gains, the future of AI development lies in multimodal capabilities and recursive self-improvement, where agents refine their own performance to solve increasingly sophisticated problems across the broader economy.