HeidelForYou

Saint's Art Design 1
Saint's Art Design 2
Saint's Art Design 3
Saint's Art Design 4
Heidel the AI ChatBot
Data Gathering from scouring the demo APU website
Code for the connections between the chatbot to APU Website
trying to apply Heidel AI Chatbot
Result of Heidel Test Script

Inspiration

The inspiration for creating Heidel stems from my personal experience back in 2023 when I was on a university hunt. I realized that universities often didn't provide clear answers to simple requirement questions, and live chat support was not always reliable due to limited availability and potential preoccupation with other inquiries. This gap in efficient and accessible communication inspired me to create an AI-powered solution that could offer reliable, 24/7 support to prospective students.

What it does

Heidel is an AI-powered chatbot designed to assist prospective university students by answering frequently asked questions related to admissions. It helps students determine the best-suited degree programs based on their interests and qualifications. as Heidel's responses are data-driven and tailored to provide accurate and relevant information, ensuring that prospective students of the universities they're applying to receive the guidance and transparent instruction they need to make informed decisions.

How we built it and Challenges we ran into

Developing Heidel wasn't without its hurdles. Communicating with the Vertex AI Agent Builder and making necessary corrections to Heidel's output were significant hurdles. The base AI model occasionally strayed off-topic, requiring continuous tuning and adjustments to ensure accuracy and relevance in responses.

Accomplishments that we're proud of

We are proud to have transformed an issue that affected me and others personally into a practical solution. The creation of Heidel represents a step forward in improving how universities can support prospective students, providing reliable and timely information that can make the their admission process questions more transparent.

What we learned

Throughout the developing of Heidel, we five gained valuable insights into data gathering and the importance of handling sensitive information responsibly. We learned about AI model tuning, website creation, and JavaScript integration for utilizing APIs effectively. such skills were crucial in bringing out Heidel to life and making it function as intended.

What's next for HeidelForYou

Looking ahead, we aim to expand Heidel's capabilities to cover more universities and their specific admission requirements. As we plan to enhance and tune better on the AI model further to improve accuracy and context understanding. Additionally, we envision integrating more interactive features, such as application process tracking. Heidel will be able to forward inquiries to the university admission team and provide updates back to the applicants on their pending questions. This will make the support system even more reliable and comprehensive. Being available 24/7, Heidel can assist millions of users simultaneously and support a wide range of languages, ensuring accessibility and convenience for prospective students from around the world.

Built With

Submitted to

Google Cloud Vertex AI Agent Builder Hackathon

Created by

I worked as the "Backend Developer," diligently procuring extensive coding and improving the chatbot's performance.

Faiyaad Tasin
I led the project from start to finish, creating the name and design for Heidel. I ensured everything ran smoothly, overseeing aspects and provide support to members when they faced challenges or idea blocks. My role involved coordinating and giving inputs to decisions, and most notably I ensuring the project stayed on track

Saint Kraugel
I worked as the "Data Acquisition Specialist," diligently procuring extensive datasets to improve the chatbot's precision and efficiency

Emmanuel Naranendhya
I handled the entire development of the chatbot, making sure it worked well and was easy to use. It was the first time I used Vertex Al and Dialogflow, which was a bit challenging, but I learned a lot. I carefully cleaned and organized the data. This took some time, but it improved the chatbot's performance. I created a set of quick replies and suggested responses. This made the chatbot more responsive and improved the user experience.

Abdullah Al Faiyaz
Private user