01
Product delivery
Turn ambiguous problems into applied AI products, then scale them across enterprise workflows.
Staff Data Scientist & Applied AI Tech Lead
Tao Cao
I build decision systems from 0-to-1 to enterprise scale, then apply the same problem-solving discipline to public AI prototypes and technical writing.
At Walmart, I lead delivery ETA and driver dispatch AI products. Outside work, I prototype agentic and quantitative tools for messy real-world decisions, and write about what I learn.
Build
0-to-1 products to enterprise scale
Science
Deep learning, optimization, simulation, agents
Production
Online and batch serving, MLOps, data pipelines
Leadership
Architecture, roadmaps, teams, stakeholders
Professional focus
Applied AI at enterprise scale
At Walmart, I lead architecture and roadmap for delivery ETA Engine and Driver Dispatch products across channels and lifecycle stages. The work spans prediction, decisioning, scalable experimentation, online and batch serving, and agentic model explainability and root-cause analysis.
These systems have delivered significant GMV, cost, reliability, and customer, driver, and associate experience impact. The work has been recognized through a KDD 2024 presentation, patent filing, and two Walmart AI Summit finalist selections.
02
Science and decisioning
Combine deep learning, optimization, simulation, conformal methods, agents, and experimentation.
03
Production engineering
Build multi-GPU training, ONNX online serving, batch inference, MLOps, and SQL/PySpark pipelines.
04
Technical leadership
Own architecture and roadmaps while guiding 5-10 DS, DE, and MLE contributors and aligning stakeholders.
Independent work
Selected prototypes
Practical tools for decisions that are hard to structure, compare, or explain.
AI Travel Planner
An agentic planning workflow for the messy middle between booking and travel. It grounds activity and food recommendations with search, then combines scoring, optimization, deterministic evaluation, and traces into an itinerary a traveler can inspect and revise.
Search grounding / Optimization / Evals / Traceability
Notes from building
Writing and talks
-
Build note
I Built an AI Travel Planner for the Messy Middle of Trip Planning
Why bounded workflows, grounding, optimization, evaluation, and traces matter in an agentic product.
-
Build note
Two Mortgage Calculators I Built to Make Prepayment Decisions Clearer
Making loan strategy tradeoffs visible through cash-flow timing, NPV, and actual payment history.
-
Build note
Turning Dull Manuals into Actionable Maintenance
A grounded multi-agent workflow for extracting, refining, and prioritizing appliance maintenance tasks.
-
Analysis
Your Heat Pump Might Be Wasting Energy Because of Your Thermostat
Quantifying a compatibility gap between variable-speed equipment and common smart thermostats.
-
Talk
Driver Search Decision-Making at KDD 2024
A TSMO Workshop talk on building large-scale AI decision systems for crowdsourced delivery.
Background
Modeling depth, production focus
I have spent more than eight years working across retail, logistics, energy, and applied research. My foundation in modeling and optimization informs how I combine deep learning, simulation, experimentation, and decisioning in production systems.
I stay close to both the technical and product work: architecture, model design, serving, evaluation, team direction, and stakeholder problem formulation.
- Ph.D.
- Mechanical Engineering, Modeling & Optimization
University of Maryland - M.S.
- Computer Science, Machine Learning
Georgia Institute of Technology - Domains
- Retail, logistics, decision systems, and energy