Build An Admission Chatbot
Gao, Liutao (2024)
2024
All rights reserved. This publication is copyrighted. You may download, display and print it for Your own personal use. Commercial use is prohibited.
Julkaisun pysyvä osoite on
https://urn.fi/URN:NBN:fi:amk-2024112028943
https://urn.fi/URN:NBN:fi:amk-2024112028943
Tiivistelmä
The thesis is an open source inspired attempt to utilize Azure OpenAI services with enterprise data to innovate the traditional complex admission system, which lacks a centralized webpage where all admission information is stored, making it challenging for applicants to access.
Additionally, the admission system does not provide personalized feedback and consulting suggestions for every applicant. As a result, some applicants fail to gain admission to the university even if they meet all criteria, leading to unhappy experiences.
To address these issues, the author will develop a ChatGPT-style chatbot. To build a functional and effective chatbot, the thesis combines Azure OpenAI Service and Azure AI Search. The thesis explains the features and structure of the application and creates a clear-structured prompt to allow the chatbot to unleash its full potential. Additionally, the thesis outlines the processes to integrate the OpenAI model with Azure Search, followed by the steps to deploy it to the Azure platform. Furthermore, the thesis conducts several tests to verify the application’s functions by asking multiple questions.
After deploying and testing the chatbot, the author concludes that the chatbot can perform unstructured tasks such as creating personalized application plans and estimating the chances of acceptance. Thus, the application can play a valuable role as an admission assistant to serve applicants effectively.
Additionally, the admission system does not provide personalized feedback and consulting suggestions for every applicant. As a result, some applicants fail to gain admission to the university even if they meet all criteria, leading to unhappy experiences.
To address these issues, the author will develop a ChatGPT-style chatbot. To build a functional and effective chatbot, the thesis combines Azure OpenAI Service and Azure AI Search. The thesis explains the features and structure of the application and creates a clear-structured prompt to allow the chatbot to unleash its full potential. Additionally, the thesis outlines the processes to integrate the OpenAI model with Azure Search, followed by the steps to deploy it to the Azure platform. Furthermore, the thesis conducts several tests to verify the application’s functions by asking multiple questions.
After deploying and testing the chatbot, the author concludes that the chatbot can perform unstructured tasks such as creating personalized application plans and estimating the chances of acceptance. Thus, the application can play a valuable role as an admission assistant to serve applicants effectively.