Vahid Azizi

About Vahid Azizi

Staff Data Scientist / Machine Learning Engineer with 6+ years of experience designing, building, and deploying large-scale ML and GenAI systems. I have led cross-functional teams as a Technical Lead, delivering ranking systems, demand forecasting pipelines, and agentic LLM applications using RAG and vector databases, driving measurable business impact across millions of users.I specialize in end-to-end ML system design — from data processing and feature engineering to scalable deployment and real-time inference. I build production-grade architecture with CI/CD, Docker, Kubernetes, MLflow, and monitoring to ensure reliability, performance, and automation.I combine deep ML expertise, GenAI innovation, system design, and strong software engineering fundamentals (DSA, modular reusable code) to create solutions that are both intelligent and highly scalable. I also mentor junior engineers and collaborate with product, engineering, and business stakeholders to align technical solutions with strategic goals.Core Technical Skills:• Machine Learning & Deep Learning: PyTorch, TensorFlow, Keras, FastAI, Scikit-learn, LightGBM, PySpark, Lightning AI• GenAI / LLM Systems: LangChain, LangGraph, LangFuse, LangSmith, Prompt Engineering, RAG, Vector DBs• MLOps & Deployment: Docker, Kubernetes, CI/CD, MLflow, Airflow, large-scale model serving, monitoring & drift detection• Data Engineering & Analysis: Python (Pandas, NumPy, SciPy), R, SQL• Cloud Platforms: AWS, GCP• Software Engineering: Data structures & algorithms, modular/reusable code• Leadership & Collaboration: Technical lead on ML/GenAI projects, stakeholder alignment, mentoring, GitHubI’m passionate about building next-generation ML and GenAI systems that combine high performance, scalability, and real business value.

Experience

1. Staff Data Scientist & Machine Learning Engineer

   Walmart Global Tech

   May 2023 — Present · Sunnyvale, CA, US

   Led the development of a production-scale demand forecasting system using PyTorch and Lightning AI, replacing the previous solution and improving bias and weighted SMAPE by ~10%, increasing inference speed by up to 8×, and reducing total cost by 10% across millions of SKUs.• Led the design and development of a multi-agent GenAI chatbot for root-cause analysis of forecast results using RAG, enabling teams and business stakeholders to diagnose issues through natural language queries and cutting analysis time from hours to minutes.• Designed and implemented a geo-based clustering framework using Weighted KMeans and H3 (Uber hexagons) to improve regional demand forecasting accuracy and scalability.• Developed Transformer-based models to generate Walmart-style category descriptions for out-of-network and new items, improving coverage and classification consistency.• Contributed to the end-to-end design and deployment of ML systems, leveraging CI/CD, MLflow, Airflow, and GCP to productionize models at scale.

Education

• K. N. Toosi University of Technology

  Master's degree, Industrial Engineering

  2011 — 2013

• Iowa State University

  Doctor of Philosophy - PhD, Major: Operations Research; Minor: Statistics

  2017 — 2021

Skills

• Powerpoint
• Microsoft Excel
• Project Management
• Microsoft Office
• English
• Python
• Matlab
• Programming
• Engineering
• Ms Project