I built what I now advise others to build.

Mugur Tolea founded Neural Kinetic in January 2026 after three decades at the intersection of technology, data, and organizational design at some of the most consequential technology organizations in the world. The conviction behind Neural Kinetic is simple: the most dangerous thing about AI advisory is advisors who have never operated inside an AI system when it fails.

At Capital One, as Vice President of Technology responsible for Customer Core Systems, Money Movement, and GenAI, Mugur led the integration that generated approximately $1B in new business impact within ten months of deployment — and spent four years maintaining zero-risk and zero-operational-incident posture across FDIC and OCC regulatory examination cycles while running a 250+ person organization. The $900M Customer Core Modernization program he led replaced a third-party core banking system with a Customer Knowledge Graph and AI-native data platform. He saw, from the inside, what it takes to operate probabilistic AI at financial services scale — including everything that goes wrong when the governance architecture is not designed before the AI is deployed.

Before Capital One, Mugur spent four years at Walmart as Vice President of Technology for Search and Pricing, architecting the search and personalization platform that served $14B in annual commerce and leading a 400–500 person organization spanning software engineering, data engineering, and data science. He built the Customer Knowledge Graph that powered GenAI-driven omnichannel personalization across Walmart’s digital properties. He spent three years at Jet.com, whose differentiated technology architecture was a key factor in its $3.3B acquisition by Walmart in 2017, and then led the post-acquisition technology integration. And he spent eight years across Audible and Amazon, building subscription commerce platforms, leading Audible’s technology integration into Amazon’s systems following acquisition, and earning the Amazon Inventor Award from Jeff Bezos in 2012 — along with two US patents.

He also served as Chief Technology Officer and Executive Vice President of Blue Foundry Bank, leading the full-stack technology transformation of a $2.5B community bank through its IPO in July 2021 — a +30% value increase on launch day.

The academic foundation beneath all of this is a PhD in Control Engineering, Neural Networks, and Fuzzy Logic from the Institut National Polytechnique de Grenoble (Grenoble INP), France — completed Magna Cum Laude in 1998. Mugur studied neural networks before they were called AI, before they ran on GPUs, and before they could be deployed at the scale the field has since reached. That foundation — mathematical, architectural, and deeply probabilistic — informs everything Neural Kinetic builds.

Neural Kinetic is the synthesis of everything that experience revealed. The advisory practice focuses on a small number of engagements at a time — organizations where the problem is real, the leadership is committed, and the window for meaningful impact is open. The published research is the intellectual contribution to the field: original frameworks, practitioner arguments, and the kind of structured thinking that only comes from having operated inside these systems at scale.

The name is deliberate. Neural: the foundational architecture of intelligence, and the field Mugur has studied since 1994. Kinetics: the science of motion and force — the translation of potential into action. Together: the conviction that intelligence without motion is just analysis. The organizations Neural Kinetic works with leave with both.

Neural Kinetic publishes original research at the intersection of AI systems design, organizational transformation, and sector-specific intelligence architecture. Every piece is a practitioner argument — built from experience, refined through advisory work, and offered to the field as a contribution to how leaders think.