Devinder Kumar
Head, ML Systems & Engineering
TD Bank (AI Platform) · Ex-Stanford Postdoc
ML Engineering Leader bridging cutting-edge research and enterprise-scale production. Currently
leading a 32-member organization at TD Bank (AI Platform), I oversee a portfolio of 75+ ML
models delivering $145M+ in annual value while reducing time-to-market by 75% through scalable
MLOps, LLMOps, and robust engineering governance. I am passionate about building
high-performance
teams and delivering AI solutions that make a difference.
Academic foundation includes a PhD in Machine Learning from the University of Waterloo and a
Postdoctoral Fellowship at Stanford University.
Core Competencies
MLOps and Platform Architecture
Designing scalable, resilient ML infrastructure serving automated decisions with high throughput and low latency.
Cross-Functional Leadership
Building, scaling, and mentoring high-performance teams of ML engineers, data scientists, and researchers.
AI Strategy and ROI
Translating complex business requirements into robust technical roadmaps that deliver scalable execution and measurable value.
Career Milestones
Head, ML Systems and Engineering
TD Bank (AI Platform)
- Leading a 32-member engineering organization (ML, Data, LLMOps, and MLOps), driving the strategic vision for enterprise-wide ML deployments.
- Architected and implemented a global enterprise MLOps target operating model, standardizing AI delivery across Canada, USA, and Singapore.
- Responsible for the operations of a portfolio of 75+ ML models in production delivering $145M annually, improving time-to-market by 75%.
Postdoctoral Fellow
Stanford University
- Advanced intersectional research at the Stanford Cognitive and Systems Neuroscience Laboratory (SCSNL), School of Medicine.
- Led cutting-edge computational modeling to solve complex challenges using large-scale deep learning paradigms.
Lead AI Scientist
NextAI (NEXT Canada)
- Promoted to Lead AI Scientist, managing a team of 4 scientists and advising C-level executives on AI R&D and product roadmaps.
- Mentored 18 distinct AI/ML startups across multiple cohorts, including the inaugural winner of the best startup award.
Deep Learning and Research Engineering
Philips Research (HQ) and UPMC-Sorbonne (LIP-6)
- Spearheaded research and engineering initiatives specializing in deep learning applications and medical oncology solutions.
- Demonstrated robust capability in foundational ML research, directly contributing to innovative global healthcare technology.
Selected Publications
Reporting radiographers’ interaction with Artificial Intelligence—How do different forms of AI feedback impact trust and decision switching?
Unsupervised Domain Adaptation in Person re-ID via k-Reciprocal Clustering and Large-Scale Heterogeneous Environment Synthesis
Insightful classification of crystal structures using deep learning
Explaining the Unexplained: A CLEAR Approach to Understanding Deep Neural Networks
Media and Impact
Yahoo Finance
"TD announces the launch of groundbreaking predictive tabular foundational model"
VICE Motherboard
"This Researcher Wants to Open the 'Black Box' of Financial AI"
EurekAlert!
"Better way found to determine the integrity of metals"
Investment Executive
"Building trust in AI"
Technology.org
"Scientists developed software to make artificial intelligence more trustworthy"
Waterloo Stories
"Reading the minds of deep learning AI systems"