HS Code classification using Agentic AI for goods import and export

Gawtham, Pathmanathan (2025) HS Code classification using Agentic AI for goods import and export. Masters thesis, Dublin, National College of Ireland.

Preview	PDF (Master of Science) Download (2MB) | Preview
Preview	PDF (Configuration Manual) Download (796kB) | Preview

Abstract

Currently within the Import/Export domain the tax code assignment of the products is handled either manually or in a semi-automated process which is time consuming and heavily dependent on domain experts. The products are also sent with improper or vague description that can cause delays in import/export. This paper explores how effectively the application of Agentic AI will classify the Harmonized System Code for import/Export of goods around the globe based on the description of the products which can improve the overall import/export process. The paper also provides a benchmark comparison by evaluating accuracy, precision, f1, recall and latency with an alternative methodology of Retrieval Augment Generation (RAG) combined with LLM. The main model that was used for implementing the Agentic AI is “gpt-04-mini”. The agentic implementation was implemented and tested with three different datasets that are similar in nature. The research provides valuable insight into the limitations and challenges of using Agentic AI for compliance and tax code classification. The proposed methodology was able attain approximately 90% accuracy on unknown HS Code descriptions and 73% accuracy for synthetic dataset with sales like product descriptions. Based on the outcomes of the research the research questions were addressed and showcased potential application of Agentic AI were identified for future research. The paper concludes by strongly suggesting that use of AI can enforce Tax Compliance efficiently.

Item Type:	Thesis (Masters)
Supervisors:	Name Email Staikopoulos, Athanasios UNSPECIFIED
Subjects:	Q Science > QH Natural history > QH301 Biology > Methods of research. Technique. Experimental biology > Data processing. Bioinformatics > Artificial intelligence Q Science > Q Science (General) > Self-organizing systems. Conscious automata > Artificial intelligence H Social Sciences > HD Industries. Land use. Labor > Business Logistics > Transportation of Goods and Trade Logistics
Divisions:	School of Computing > Master of Science in Data Analytics
Depositing User:	Ciara O'Brien
Date Deposited:	07 Sep 2026 10:50
Last Modified:	07 Sep 2026 10:50
URI:	https://norma.ncirl.ie/id/eprint/9860

Actions (login required)

View Item