Date of Graduation
5-2026
Document Type
Thesis
Degree Name
Bachelor of Science in Data Science
Degree Level
Undergraduate
Department
Data Science
Advisor/Mentor
Dr. Karl Schubert
Committee Member
Dr. Eric Specking
Second Committee Member
James McGinley
Third Committee Member
Gordon Morrisette
Abstract
This thesis documents the design, development, and deployment of Spark Ask, an AI-powered data analysis tool built during an internship at Spark Strategy, a Walmart representative group based in Bentonville, Arkansas. Non-technical business analysts at the firm were unable to run custom data reports without developer assistance, slowing decision-making across large volumes of point-of-sale, inventory, and supply chain data. Spark Ask addresses this by accepting plain-English prompts, generating the corresponding Python code via a large language model, executing that code automatically, and returning a CSV report with no programming required. The system consists of a Flask REST API hosted on AWS Lightsail, a PySide6 desktop application, and an OpenAI large language model configured with a detailed data dictionary and YAML-based instructions to handle Walmart-specific data structures. Results show that business staff can now generate accurate reports on demand across seven standard vendor datasets, significantly reducing developer involvement in routine data requests.
Keywords
Artificial intelligence, natural language processing, Python code generation, large language model, Flask REST API, AWS Lightsail, PySide6, data analysis automation, business intelligence, point-of-sale data, supply chain data, inventory management, Walmart vendor data, OpenAI, data dictionary, plain-English querying, CSV reporting, developer workflow reduction
Citation
Mendez, A. F. (2026). Spark Ask: Leveraging Large Language Models for Natural Language Driven Data Analytics in Walmart Retail Vendor Management. Data Science Undergraduate Honors Theses Retrieved from https://scholarworks.uark.edu/dtscuht/37
