Redesigning Workflows: Uber’s Unlikely AI Breakthrough

The phrase “tokenmaxxing” has become synonymous with the perils of unchecked AI enthusiasm, but Praveen Neppalli Naga, Uber’s CTO, is pushing back with a bold new strategy. Rather than speeding up individual tasks with AI, Naga argues that real productivity gains come from radically rethinking workflows.

Uber’s numbers are impressive: financial planning was cut from 15 hours to 30 minutes, and financial reports reduced from two days to 10 minutes. Marketing quality checks shrunk from two weeks to under an hour. But it’s not just about efficiency; Naga is pointing to a fundamental shift in how companies approach AI implementation.

Uber has abandoned the “magic bullet” approach, where AI was seen as a pricey experiment or a quick fix. Instead, the company is embedding its top engineers directly into corporate processes. This involves pairing skilled engineers with domain experts to tackle complex workflows head-on. The result: short-term teams called Agentic Pods that shadow employees for days and then rapidly build and test AI-powered tools.

By acknowledging the limitations of traditional AI implementation, Uber is forcing us to confront a harsh reality: AI’s potential lies not in individual tasks but in fundamentally transforming how work gets done. This means more than just automating repetitive processes; it requires challenging entrenched practices and rethinking what we consider “work” in the first place.

Uber’s experiment has sparked questions about whether other companies will follow suit. The finance sector, with its labyrinthine approval processes and outdated software, seems ripe for disruption. Marketing and HR departments could also benefit from Naga’s approach.

However, there are risks involved. By emphasizing the importance of workflow redesign, Uber is drawing attention to a fundamental flaw in AI implementation: we often try to fit new technology into existing structures rather than letting it reshape our processes entirely. This might lead companies down a path of incremental change, incrementally improving individual tasks without ever tackling deeper issues.

As Naga continues to push the boundaries of what’s possible with AI, one thing is clear: Uber’s CTO has sparked a much-needed conversation about how we approach AI implementation. The implications for the future of work itself are significant.