Uber Reinvents AI Internally After Tokenmaxxing Panic

⚡

Key Takeaways

1Praveen Neppalli Naga, CTO of Uber, returns with a new AI strategy following the tokenmaxxing panic.

2Uber is deploying "Agentic Pods" to observe and optimize internal workflows with AI.

3Uber's financial and marketing processes have been drastically accelerated thanks to this approach.

💡Why it matters — This strategy could transform Uber's internal productivity, potentially influencing other companies to adopt this model.

A New Vision of AI at Uber

Praveen Neppalli Naga, Uber's Chief Technology Officer, is back in the spotlight with an innovative approach to artificial intelligence. Earlier this year, he caused a stir by announcing that Uber had already consumed its 2026 budget for Anthropic's Claude Code, an event that sparked panic around the concept of tokenmaxxing. Today, he proposes a different vision, focused on optimizing internal processes through AI.

"Agentic Pods": Immersion in Employees' Daily Lives

Uber has launched a bold initiative by sending its artificial intelligence experts into various departments such as finance, legal, marketing, customer support, human resources, and procurement. The goal is to closely observe the daily operations of employees before developing AI tools tailored to their specific needs. These teams, called "Agentic Pods," immerse themselves for two weeks in employees' activities, spending the first few days understanding every step of the workflows, then quickly designing and testing software solutions to improve them.

Impressive Results

The results achieved through this method are spectacular. For example, a financial planning process that initially took 15 hours has been reduced to just 30 minutes. Similarly, financial reporting, which required two days, can now be completed in 10 minutes. Quality checks in marketing, previously spread over two weeks, are now done in under an hour.

Rethinking Workflows with AI

According to Naga, the greatest benefits lie not only in speeding up tasks but in completely rethinking workflows around AI. This includes eliminating unnecessary approvals, replacing outdated software, and aiding in rapid decision-making. This approach allows for a deeper and more effective transformation of processes.

A New Trend in Silicon Valley

Uber's strategy has been compared by several tech leaders to that of "forward-deployed engineers" in Silicon Valley, who are typically sent to clients to improve their systems. However, Uber is innovating by deploying these experts within its own structure, seeking to uncover new productivity opportunities. Peter Wilczynski, product director at Vantortech, even joked by referring to this role as "Rearward Deployed Engineer," highlighting the deep integration of engineers into internal operations to reinvent workflows through AI.