Bret Taylor (Sierra CEO & OpenAI Chair) on AI Agents, SaaS, and the AI Bubble

The Rise of AI Agents

• 💡 Sierra builds AI agents for customer experience, particularly in financial services, replacing traditional IVR systems and handling tasks like fraud reports and password recovery.
• 🚀 These agents are becoming a primary digital touchpoint, as important as mobile apps or websites, offering immediate, 24/7 service at a fraction of the cost of human-led interactions.
• ✅ AI agents demonstrate significant ROI, with examples like Rocket Mortgage seeing 3x higher transaction completion and Ramp achieving over 90% self-service resolution rates.

Future of Customer Experience with AI

• 🧠 The future of AI agents is multimodal, enabling seamless interactions across voice and digital chat.
• 🎯 Agents will be personalized, understanding user context, preferences, and intent to offer tailored support.
• 🌱 A key trend is proactivity, where agents solve problems or offer solutions before customers even realize an issue exists, like notifying travelers about roaming charges.

Outcomes-Based Pricing Model

• 💰 Sierra champions outcomes-based pricing, where payment is contingent on the successful resolution of a customer case by the AI agent.
• 🤝 This model fosters true partnership between software vendors and customers, aligning incentives and providing clear, measurable ROI.
• 📈 It represents a seismic shift from traditional subscription models, creating accountability and transparency in software value.

Sierra's Competitive Edge

• ✨ Sierra's agents are highly effective, evidenced by high customer satisfaction scores and impressive resolution rates in real-world applications.
• 🔒 The company specializes in serving regulated markets like financial services and healthcare, navigating complex compliance requirements.
• 🔑 Their unique outcomes-based business model differentiates them by ensuring customers only pay for demonstrated value.

AI Market Dynamics and Moats

• 🌐 The AI era is structured with model companies (like OpenAI) providing foundational infrastructure and application-layer agents addressing specific workflows.
• 🏰 Future moats will increasingly be found in product differentiation and specialized solutions for key departments, rather than just underlying technology.
• ⚠️ The current AI landscape is a "healthy bubble," where the underlying technological transformation is real and significant, but many companies will fail, similar to the 1999 dot-com era.