Why AI Will Transform Customer Experience: Cresta CEO Ping Wu and Sequoia’s Doug Leone

The Evolving Contact Center Landscape

• 💡 The contact center market is vast, encompassing omnichannel interactions (calls, emails, chats) and employing 17-20 million human agents.
• 🎯 Historically, it's been a "gnarly" legacy industry with high agent attrition (35-40%) and dissatisfaction among customers, agents, and businesses.
• 📈 Technology has evolved from IVR systems to NLP and pre-transformer models, with Large Language Models (LLMs) now fundamentally changing conversation understanding and automation.

AI's Transformative Impact

• 🧠 While some predict 100% automation, experts believe the transition for large enterprises will be slower and more nuanced, with full humanless contact centers unlikely in the near future.
• 🚀 The key metric is not the final automation percentage but the speed of adoption, which varies significantly based on product complexity and IT infrastructure.
• ✨ AI can improve customer experience by fixing root causes of issues, automating low-value interactions, and enabling new, previously unaffordable interactions like asynchronous app communication.

Cresta's Dual Approach to AI

• 🛠️ Cresta employs a dual strategy: agent assist tools to boost human productivity and autonomous AI agents for direct customer interaction.
• 📊 The agent assist product is crucial as it provides real-time conversation data, which then serves as a training base for developing more effective autonomous agents.
• 🔑 Overcoming deployment challenges involves addressing legacy IT systems that lack APIs and building vertical integration from "stake to sizzle" to deliver immediate value.

Building AI Companies with Speed

• ⚡ Building AI companies requires extreme speed and a focus on core business fundamentals: exceptional founders, world-class engineers, and effective sales and marketing.
• ⛰️ Leaders must identify and remove "rocks" (e.g., funding, management experience) that hinder growth, constantly pushing for faster, linear revenue ramps.
• 📈 The AI wave is considered an "Industrial Revolution 2.0", far larger than previous internet or mobile waves, characterized by non-linear improvements and constant surprises from underlying model capabilities.

Technical Challenges and Solutions

• 🔬 Cresta's voice AI agent orchestrates multiple models (speech-to-text, noise cancellation, foundation models, TTS, guardrails) bidirectionally, achieving sub-800 millisecond latency.
• ⚙️ Deploying AI in production faces significant hurdles, including real-time audio access from on-premise systems, handling complex call flows (transfers, third parties, long durations), and ensuring PII compliance and data residency.
• ✅ Conversational flow is managed by giving LLMs goals and tools, using testing, simulation, and guardrails to ensure deterministic and safe behavior, rather than rigid, script-based interactions.

The Future of Customer Experience

• 🔮 AI will eventually "disappear" into workflows, making interactions seamless and personalized, where users may not even realize they are talking to an AI.
• 💬 The vision is a continuous customer journey, where AI agents provide a consistent, personalized experience across all phases (sales, service), eliminating the "multiple personality" feel of current businesses.
• 🌱 This shift from a scarcity to an abundance mindset will enable entirely new customer experiences, such as talking directly to airline apps for asynchronous task completion.