AivexaNewsSearch
AI news for builders and product teamsChecked every hour

The Foundations of Scalable AI

Collected Oct 22, 2026

MIT Technology Review's EmTech Future conference is hosting a session titled "The Foundations of Scalable AI," presented by Vanguard, featuring Jing Wang, principal of Vanguard's Center for AI and Insights (CAI). The session focuses on the premise that a scalable AI system depends on a strong foundation, with data, engineering, platforms, operating models and governance converging to work as one. It also examines how to build that foundation inside a highly regulated environment and how a responsible, trustworthy implementation can turn emerging capabilities, including agentic systems, into lasting value for investors. No date or release details are attached to the listing beyond its appearance on the publication's AI channel.

The agenda cuts across several layers of an AI stack at once. The framing treats data, engineering and platforms as the technical base, then adds operating models and governance as equally load-bearing parts. In practice, this is an argument that a model alone is not a system: the surrounding processes, controls and ownership structures decide whether an AI capability can scale. For a firm operating under financial regulation, governance is not an afterthought but a design constraint, which is why the session pairs scalability with trustworthiness rather than treating them as a trade-off.

Agentic systems are called out as one of the emerging capabilities under discussion. The term usually describes AI that can take multi-step actions, invoke tools and pursue a goal with limited step-by-step human direction, rather than only answering a single prompt. Because the stated theme is lasting value for investors, the implied question is how such capabilities can be deployed in a way that survives compliance scrutiny and produces benefits that hold up over time. The listing does not detail specific products, benchmarks or rollout plans.

Wang's background is largely organizational and quantitative. She leads CAI, which advances Vanguard's AI, machine learning, generative AI and behavioral science capabilities. Her team drives enterprise-wide innovation, research and delivery, with the aim of translating emerging technologies into trusted solutions that improve client experiences and investor outcomes and create measurable business value. The inclusion of behavioral science alongside machine learning suggests attention to how people and clients actually interact with these systems, not only to model performance.

Before her current role, Wang held leadership positions across several Vanguard divisions, including Financial Services, Workplace Solutions, International and Corporate Services, leading teams and enterprise initiatives tied to strategic priorities and client outcomes. She joined Vanguard in 2009 after several years as a principal at Boston Consulting Group. She holds a B.S. in Physics from the University of Science and Technology of China, a Ph.D. from the University of Pennsylvania and an M.B.A. in finance and marketing from Wharton, plus Series 7, 24 and 65 licenses. That mix of a research doctorate, consulting and registered-securities credentials fits the session's dual emphasis on emerging technology and regulated delivery.

This reads as a practitioner's account rather than a product announcement. The audience is likely to come away with a framework for sequencing foundation work: consolidating data and platform layers, establishing operating ownership, and encoding governance before scaling to more autonomous capabilities. A likely trade-off is speed, since governance-heavy pipelines typically slow the path from prototype to production, and the session is positioned to argue that this cost is what makes durable value possible. Whether the talk offers concrete architectures, failure cases or metrics is not indicated.

Why it matters: Teams building AI in finance and other regulated sectors face the same question this session poses, namely whether their data, platform and governance layers can carry more autonomous systems without breaking compliance or client trust. For developers, the emphasis on operating models alongside engineering suggests that production readiness will be judged on controls and ownership as much as model quality, an inference drawn from the session's stated themes rather than a reported result. Product leaders may find the investor-outcomes framing useful when justifying AI investment internally.

Read at MIT Technology Review · AI

Based on reporting from the original publisher. Visit the source for full context and later updates.

Publisher excerpt

A scalable AI system requires a strong foundation. Data, engineering, platforms, operating models, and governance must converge to work as one. This session from MIT Technology Review’s EmTech Future conference explores what it takes to build that foundation in a highly regulated environment and how a responsible, trustworthy implementation can turn emerging capabilities, including agentic…