Sandeep Mahajan is Director of Data Science at Walmart, where he works at the intersection of operations, data, and people to deliver AI and analytics that change how work gets done at scale. His career has been shaped by the belief that technology only creates value when it is grounded in real workflows and adopted on the front line.
Sandeep began his career in large-scale operations and supply chains, with early experience implementing SAP and RFID systems. These roles taught him that data quality, governance, and user adoption matter far more than elegant models alone. He carried these lessons into product and data leadership roles at Sears, Amazon, and Walmart, where he repeatedly saw that analytics initiatives succeed or fail based on how well they align with day-to-day decision-making.
A defining turning point came at Walmart, where he shifted focus from building isolated machine learning solutions to leading decision-oriented, agentic systems. By reframing the problem from “what model can we build?” to “what decision are we improving, for whom, and under what constraints,” he helped create autonomous store agents and workforce intelligence platforms that delivered measurable enterprise impact.
As a leader, Sandeep balances ambition with pragmatism, combining rapid experimentation with strong governance and accountability. He is known for a servant-leadership approach, removing friction, building trust across disciplines, and keeping humans at the center of AI systems. His perspective today is rooted in the belief that the future of data and AI lies in responsible, decision-centric systems that empower people and scale with integrity.
As a data and AI leader, which traits and skills do you think matter most, and which of those have been most influential for you in your current position?
“In my experience, the most effective data and AI leaders combine decision clarity, systems thinking, and human judgment. Technical fluency matters, but it’s rarely the differentiator. What truly determines impact is the ability to frame the right problems and design intelligence that fits real-world constraints.
“The trait that has mattered most in my organization is decision ownership. Data leaders are often asked to “provide insights,” but meaningful impact comes from owning outcomes – being accountable for whether a decision actually improves performance. This mindset shifts teams from optimizing models in isolation to designing end-to-end systems that influence behavior.
“A second critical skill is cross-functional influence. Data and AI sit at the intersection of engineering, operations, risk, and frontline users. The ability to build trust across those groups by listening deeply, translating trade-offs, and aligning incentives has been essential to scaling solutions beyond pilots.
“Finally, pragmatic judgment has been influential. Knowing when to experiment and when to standardize; when to automate and when to keep a human in the loop. In my organization, trust was built not by pushing AI everywhere, but by being explicit about limits, failure modes, and escalation paths.
“The leaders who succeed are those who treat data and AI not as tools to showcase sophistication, but as systems that must earn trust, deliver value, and fit the people who rely on them.”
Reflecting on your career, what is one non-traditional piece of advice (outside of technical skills) you would give to an aspiring data or AI leader aiming for the C-suite?
“A non-traditional piece of advice I would give is to spend time owning uncomfortable, operational problems that aren’t data-ready. Early in my career, the experiences that shaped me most weren’t elegant analytics projects, but messy situations involving broken processes, unclear ownership, and frontline resistance. Sitting with that discomfort taught me how decisions are really made and why many data initiatives fail.
“For aspiring C-suite leaders, credibility comes less from technical brilliance and more from being trusted to navigate ambiguity, trade-offs, and human dynamics. Seek roles where success isn’t guaranteed and where influence matters more than authority. Learn how to say ‘no’, how to explain uncertainty, and how to protect teams when experiments fail.
“The best data and AI leaders I’ve worked with are those who understand that leadership isn’t about having the smartest answer, it’s about creating conditions where the organization can make better decisions repeatedly, even when the data is incomplete and the path forward isn’t obvious.”