Kamyab Nazari

Project

Customer Service Classifier AI

This research initiative explores Large Language Models' effectiveness in categorizing banking customer inquiries. The project employs experimental approaches with prompt-engineering and model fine-tuning techniques to strengthen customer service operations.

Key Features

BANKING77 dataset — a vetted collection of labeled banking inquiries
Zero-Shot and Few-Shot prompting techniques
Model fine-tuning using specific dataset segments
Standardized ML metrics for objective performance assessment
Cost-effectiveness analysis of implemented methods

Repository

View on GitHub

Technologies

AI
GPT-3.5 Turbo
Prompt Engineering
Fine-tuning
Python