Pedram Farghadani

Pedram Farghadani

Data Scientist focused on optimization, forecasting, and AI decision support

About

I am a Data Scientist at Yeji Data Lab with a PhD in Operations Research from Concordia University and CIRRELT. I build machine learning, optimization, and AI decision-support systems for complex problems in forecasting, logistics, and healthcare.

What I Build

Forecasting pipelines, optimization models, reinforcement learning systems, and LLM-powered tools that help teams make better decisions under uncertainty.

How I Work

I combine rigorous modeling, practical implementation, and clear communication to turn research-grade methods into useful operational tools.

Current Focus

  • Optimization and decision intelligence for operational planning
  • Demand forecasting and feature-rich predictive modeling
  • LLM, RAG, and prompt-based systems for technical workflows
  • Research-to-production translation for real-world decision support

Selected Highlights

  • Built demand forecasting models on large-scale transactional sales data using XGBoost and LightGBM.
  • Developed applications with LangChain, Azure OpenAI, and FAISS for retrieval, question answering, and optimization-assisted planning.
  • Published work in Computers & Operations Research, and International Journal of Production Research.
  • Presented research at venues including CORS, Optimization Days, and the International Conference on Stochastic Programming.

Background

My training sits at the intersection of operations research, machine learning, and applied analytics. I completed my PhD in Operations Research at Concordia University, supervised by Professors Hossein Hashemi Doulabi, Walter Rei, and Michel Gendreau. I previously earned an MSc in Logistics and Supply Chain Management from Amirkabir University of Technology, under the supervision of Professor Behrooz Karimi, and a BSc in Industrial Engineering from Isfahan University of Technology.

For a fast overview, see my Projects, Professional Experience, and Publications.