Saviz Saei, Ph.D.

Research Scientist | Social Science Research Center, MSSTATE

• Launched a full-stack student Risk Assessment System: a modern full-stack web application to track student risk assessments with RAG model Chatbot based on Role. [Demo]
• Conducted statistical analyses (t-tests, chi-square, Mann-Whitney U, regression modeling) on healthcare access and eHEALS survey data using Python, R, and SPSS.
• Designed an AI RAG model and service-matching agent with LLaMA & OpenAI. [Paper - ready to publish]

Research Scientist | Social Science Research Center, MSSTATE

• Conducted health behavior analytics based on eHEALS, such as principal component analysis, t-tests, chi-square, Mann-Whitney U, regression model to inform strategy for digital health among older adults using Python, R, and SPSS.
• Designed an AI RAG model regarding service-matching agent with LLaMA & OpenAI. [Paper - ready to publish]

Machine Learning Engineer | Archer Daniels Midland (ADM)

• Re-architected LLM SAP Report: rebuilt SAP-generated reporting with Python/JavaScript; delivered containerized services using FastAPI and Faiss (vector DB), integrated with APIM and React, and deployed to Azure, reducing reporting cost by ~30% and increasing throughput.
• GenAI Q&A (RAG) Compass Report: delivered an Azure-hosted RAG system (Python, Dash) leveraging embedding-based cosine similarity to improve findability and self-service analytics.
• GenAI Q&A HR Recommendation System: built an embedded similarity search–based candidate matching engine on Azure, cutting screening time by ~40%.
• Cost Optimization: monitored cost in Azure and adjusted the infrastructure to optimize spending.

Research Scientist/Intern | Social Science Research Center

• Analyzed social media data for an NSF-funded research project using Pandas, NumPy, NLP (NLTK, NRCLex, regex).
• Built interactive data visualizations with JS & Matplotlib. [Paper in review]

Research Assistant | Mississippi State University

• Conducted an in-depth study on disaster resilience across engineering, ecology, and social sciences to detect vulnerabilities and guide mitigation strategies.
• Scenario-based network resilience optimization combining vulnerability assessments with traffic flow using Gurobi.
• Ranked nodes via different network metrics to build disruption scenarios for the optimization model.
• Developed strategic board game AI with machine learning and neural networks; integrated A* search and deep reinforcement learning to improve decision quality.
[AI Project]
• Researched on neural network in geometric metrics,policy gradients, Q learning, and actor-critic methods to advance sequential decision making.
[Neural Network in Geometric Metrics Project]

Research & Teaching Assistant | Ohio University

• Performed regression analysis and SQL preprocessing with IBM on human trafficking data.
• Served as TA for business analytics (SQL, Access, Excel); guided students in labs.

Project Manager & Data Scientist | Golrang System Company-IT

• Managed portfolios across data analytics, sales, and web development; streamlined task management and team collaboration, improving delivery timelines by ~25%.
• Applied agile methodologies to optimize workflows, strengthen cross-functional alignment, and accelerate releases.
• Developed and deployed analytics to enhance decision making, customer engagement, and boost sales performance (~15%).
• Translated business requirements into technical roadmaps with senior leadership; ensured measurable value and adoption.
• Enabled large-scale IT and data initiatives that increased operational efficiency and advanced digital transformation across Golrang Industrial Group.

Network resilience

Vulnerability analysis, restoration ordering, and scenario-based traffic assignment on disrupted networks.

Optimization under uncertainty

Two-stage stochastic programming and robust formulations; heuristics for routing and allocation problems.

Applied Machine Learning

End-to-end ML systems in Python, TensorFlow, PyTorch; pipelines and evaluation.