Ruba Borno leads AWS’s Global Specialists and Partners organization, working with companies that use Amazon Web Services to turn AI advances into business impact.
To help make the case, she’s also turned her own AWS Marketplace engineering organization into a unit where humans define the requirements and artificial intelligence generates the code, backed by autonomous agents that triage incidents and deliver fixes without waiting for a prompt. As a result, the group increased its shipping throughput by 88%, improved time to production by 21%, and maintained steady rollback rates. More than 76% of the code that goes into production is now AI-assisted. Borno has set an unambiguous internal tone: “We must use AI every day in nearly everything we do,” as she put it in a talk at McKinsey.
Under her leadership, AWS Partner Central introduced AI agents that qualify sales opportunities on arrival, deliver tailored guidance, and route each deal to the right people. “For partners and customers, that means deals move faster and the right resources get matched sooner,” she says, adding that more than 1,000 partners have adopted the agents since launch.
She also oversaw a reinvention of AWS Marketplace, which now features an agent mode that conversationally guides customers to offerings, as well as “express private offers” that let customers get personalized pricing in minutes rather than through lengthy sales discussions.
The first key to getting worker buy-in and finding success with AI pilot programs, Borno says, is making sure that the AI tools have access to the right contextual data. This means building the plumbing (infrastructure) to connect the right data, label and organize it so it’s easy to find, and set controls to ensure only the right people or agents have access, she explains. “You don’t have to unify your entire data warehouse on day one. Work backwards from a specific use case, get that data right, and scale from there.” Without the right data, she adds, the AI models won’t have the ground truth data needed to generate useful outputs, and workers will soon lose trust.
Borno says businesses also need to think differently about measuring the success of AI programs. Most companies, she notes, try to measure AI ROI against existing processes and existing budgets, and that’s why so many pilots stall. “The greatest value of AI doesn’t come from doing the same work cheaper,” she says. “It comes from doing work that wasn’t possible before.”