Uber is expanding its use of artificial intelligence beyond software development by embedding teams of AI specialists directly into business departments to redesign workflows and build AI agents capable of automating complex operational tasks.
The initiative, known internally as “Agentic Pods,” reflects a growing shift among large technology companies from using AI primarily as a coding assistant to deploying autonomous AI systems that can perform multi-step business processes traditionally handled by employees.
Praveen Neppalli Naga, Uber’s technology chief, said in a post on X that the company assigned 30 of its most AI-proficient engineers to work alongside employees in departments including finance, legal and human resources.
Register for Tekedia Mini-MBA edition 20 (June 8 – Sept 5, 2026).
Register for Tekedia AI in Business Masterclass.
Join Tekedia Capital Syndicate and co-invest in great global startups.
Rather than relying on process documentation, the engineers spent two weeks observing how teams actually performed their day-to-day work before building AI agents tailored to those tasks.
“You can’t automate them effectively by looking at process diagrams or documentation,” Naga wrote. “You have to understand how the work actually gets done.”
Over the past two months, Uber has completed 16 Agentic Pods, each focused on identifying repetitive, time-consuming work that could be redesigned with AI.
AI Agents Target Complex Business Processes
Unlike conventional automation software that follows predefined rules, Uber’s AI agents are designed to interact with multiple internal systems, gather information, analyse data and complete workflows that previously required human intervention.
Many of the projects involved tasks that employees performed manually across several software platforms. One example cited by Naga was the preparation of financial pacing reports, which previously required staff to collect information from multiple systems over two days.
With AI agents handling much of the process, those reports can now be generated in approximately 10 minutes.
Another workflow involved allocating capital across the roughly 150 cities where Uber operates. According to Naga, a task that previously required about 15 hours of manual analysis now takes around 30 minutes using AI agents.
Uber’s approach also highlights the growing importance of a relatively new role within the technology industry. Instead of remaining within engineering departments, AI specialists are increasingly working directly with business teams to understand operational challenges before designing AI-powered solutions.
The so-called forward-deployed engineers have become one of the few technology hiring categories to remain active even as broader layoffs continue across the industry. These engineers typically work closely with customers or internal business units, translating operational needs into AI applications rather than focusing solely on software development.
Uber has effectively adopted that model internally by embedding engineers within its own departments.
The strategy is seen as a recognition that many business processes rely on informal knowledge and practical experience that cannot easily be captured in manuals or workflow diagrams. Observing employees perform their work often reveals shortcuts, exceptions and decision-making patterns that AI systems need to replicate effectively.
AI Spending Under Scrutiny
Uber’s expansion of AI agents comes as technology companies continue investing heavily in artificial intelligence while facing growing pressure from investors to demonstrate measurable returns.
Like many large technology firms, Uber has significantly increased spending on AI infrastructure, software tools and model access.
Earlier this year, Naga told The Information that Uber exhausted its annual budget for Anthropic’s Claude Code AI assistant well before the end of the year, underscoring the rapid pace of the company’s AI adoption.
However, senior executives have acknowledged that translating those investments into customer-facing products has proved more challenging.
In May, Uber Chief Operating Officer Andrew Macdonald said on a podcast that the company was finding it increasingly difficult to justify the scale of its AI spending. Although AI has improved internal productivity, Macdonald said the investments have not yet resulted in a proportional increase in useful consumer features for riders or drivers.
However, Uber’s Agentic Pods underpins one of the fastest-growing areas of enterprise AI: the deployment of autonomous agents capable of carrying out sequences of tasks with limited human supervision. Unlike traditional chatbots, AI agents can retrieve information, make decisions based on predefined objectives, interact with enterprise software, and complete end-to-end workflows.
Many technology companies see such systems as the next stage in workplace automation, potentially reshaping functions ranging from finance and legal operations to customer service and supply chain management.
For Uber, the focus appears to be on improving internal efficiency before expanding AI capabilities into consumer products. Naga said the company intends to scale the initiative by establishing a dedicated team focused exclusively on identifying business processes that can be redesigned with AI.
“We’re now forming a dedicated team to scale this further and go deeper,” he said.
“They’ll deeply understand the work, redesign it from the ground up, and use AI to fundamentally change how the business operates.”
The initiative suggests Uber views AI not simply as a tool for incremental productivity gains, but as a means of fundamentally restructuring how work is performed across the company.