Navan Cofounder’s Bold AI Bet Pays Off
The Challenge That Changed Everything
Ariel Cohen, cofounder and CEO of corporate travel startup Navan, issued a radical ultimatum to his company’s new agentic AI assistant: either autonomously complete his entire international business trip—from flights to meetings—or be scrapped. What followed wasn’t just an organizational experiment but a pivotal trial of how far generative AI has come from being a chatbot to becoming a business-grade autonomous agent. Cohen’s gamble was born from both urgency and belief; as the AI revolution accelerates across Silicon Valley, Cohen saw no room for half-measures. If Navan’s AI couldn’t handle the complex, real-world task of planning and executing a full trip, it wouldn’t be fit for the company’s future.
A Surprising Victory for Agentic AI
To Cohen’s astonishment, the AI passed. Dubbed “Ava,” Navan’s agent didn’t just book flights and hotels—it optimized schedules, accounted for delays, and adjusted in real time without human interference. The success sent ripples through Navan’s leadership, galvanizing the team to prioritize AI development more aggressively and reimagine how autonomous systems could transform business operations. While many companies still tread cautiously around giving AI full autonomy, Navan’s proactive testing under pressure may hint at a new era of enterprise AI adoption—even if the stakes are just one CEO’s travel calendar.
