This is a profile image of Stephen Xu

Stephen Xu

PartnerToronto

Transforms our clients’ core business processes using AI, gen AI, and agentic AI; builds code assets to accelerate and scale deployments

Stephen leads McKinsey’s gen AI and R&D efforts within QuantumBlack Labs, the software development and R&D hub of QuantumBlack, McKinsey’s AI arm. He also leads QuantumBlack’s machine learning R&D program.

His product teams have supported 150+ gen AI product builds globally across many sectors and functions. These teams also convene a community of 200+ technical practitioners (i.e., data scientists, machine learning engineers, and cloud architects) to share best-practice techniques and exchange code.

Stephen is also a lead technical adviser for McKinsey’s open ecosystem of gen AI partnerships, providing counsel on how we can bring the best of the tech industry to our clients. These organizations span vertical software as a service (SaaS) players, horizontal tooling players, and foundation model providers.

Examples of Stephen’s recent client work include the following:

  • led multiyear agentic AI product build for a legal services provider, delivering the industry’s first agentic AI solution to the market and bending the cost curve (reduced end-to-end handling times by 4X and halved costs)
  • led rollout of sales agents powered by Agentforce for a B2C player, doubling AI agent conversion rates with state transition analysis
  • helped design a gen AI platform for a leading North American telco, architecting shared services and components to accelerate gen AI build efforts across the enterprise
  • led design and implementation of a gen AI governance and operating model for a leading North American financial services player, including architecture, model risk management, and portfolio governance
  • built a supply chain digital twin for an aerospace and defense company with demand forecasting and inventory optimization capabilities, driving reduction in on-hand inventory and improved customer SLAs

In addition to his client work, Stephen is a professional pianist, holding the highest practical examination achievement with the Royal Conservatory of Music; a mathlete, placing top 50 in Canada’s U18 Mathematical Olympiad Invitational; and an amateur go-kart racer.

PUBLISHED WORK

One year of agentic AI: Six lessons from the people doing the work,” McKinsey & Company, September 2025

From promising to productive: Real results from gen AI in services,” McKinsey & Company, August 2024

Why agents are the next frontier of generative AI,” McKinsey & Company, July 2024

Moving past gen AI’s honeymoon phase: Seven hard truths for CIOs to get from pilot to scale,” McKinsey & Company, May 2024

PAST EXPERIENCE

M-KOPA
Data scientist

NVIDIA
Software engineer

EDUCATION

University of Toronto
BSc, engineering science