AWS Builder Center

Building a Club Member Portal with AI for AWS Student Builder Groups

Description We built a simple Club Member Portal for AWS Student Builder Groups where members can sign up, log in, reset their password, and ask questions through an AI chatbot. The chatbot uses the provided club documents to answer questions and shows the source of the information.

Introduction
As part of the AWS Student Builder Groups hackathon, our team worked on building a Club Member Portal for students.
The main purpose of our project is to make it easier for club members to access useful information and get answers to their questions in one place.
A new member may have questions such as when the next workshop is, how to publish an article on Builder Center, or how to get started with AWS. Instead of searching through different documents, the member can ask the chatbot directly.
What We Built
We built a Club Member Portal with the following features:
- Sign up using email and password - Login for existing members - Forgot password / password reset - A chat page available only to logged-in members - Club documents used as the knowledge source - AI-based answers from the provided documents - Source information shown with the answers - A fallback response when the information cannot be found
The project is based on the starter documents provided for the hackathon, which contain information about onboarding, AWS account setup, Builder Center publishing, Amazon Bedrock, hackathon rules, workshops, Lambda patterns, and Student Builder Groups.
Why We Built This
When a student joins a Student Builder Group, there can be a lot of information to go through.
For example, a member might want to know:
- How do I publish an article on Builder Center? - When is the next workshop? - How do I get started with Bedrock? - What are the hackathon rules?
We wanted to make this information easier to access.
With our portal, members can log in and ask their questions through the chat interface instead of manually searching through different documents.
How the Chatbot Works
The chatbot is designed to use the club documents as its source of information.
When a member asks a question, the system searches the available documents for relevant information and uses it to provide an answer.
We also wanted the user to know where the answer came from. Therefore, the chatbot displays the source document or section along with the answer.
This makes it easier for members to verify the information.
Safe AI Responses
One of the important parts of our project is handling questions that are not covered by the documents.
Instead of guessing or making up an answer, the chatbot gives a fallback response when it cannot find the required information.
The member is then directed to the campus AWS Student Builder contact for further help.
This follows the hackathon requirement that the AI should answer only from the club documents and should not invent AWS pricing, limits, or policies.
Member Authentication
The portal includes a simple authentication flow.
A new member can create an account using an email address and password.
Existing members can log in using their credentials.
We also included a forgot-password option so that a member can request a password reset.
The chat page is available only after the member has logged in.
Project Flow
The basic flow of our application is:
1. A member opens the Club Member Portal. 2. A new member creates an account. 3. The member logs in. 4. The member opens the chat page. 5. The member asks a question. 6. The system searches the available club documents. 7. The chatbot provides an answer based on the documents. 8. The source of the answer is displayed. 9. If the information is not available, the chatbot provides the fallback response.
Screenshots
Sign Up
[Add your Sign Up screenshot here]
This screen allows a new member to create an account using their email and password.
Login
[Add your Login screenshot here]
This screen allows an existing member to log in to the portal.
Forgot Password
[Add your Forgot Password screenshot here]
This screen allows a member to start the password-reset process.
Chat Page
[Add your Chat Page screenshot here]
After logging in, members can use the chat page to ask questions about the club and the information available in the starter documents.
Answer With Source
[Add your Chat Answer screenshot here]
The chatbot provides an answer and shows the source document or section used for the answer.
Fallback Response
[Add your Fallback Response screenshot here]
When the required information is not available in the documents, the chatbot does not guess. It provides a helpful fallback and directs the member to the campus AWS Student Builder contact.
AWS Deployment Plan
For our prototype, the application can be demonstrated locally.
If we move the project to AWS, we would use different AWS services for different parts of the system.
Amazon Cognito can be used for member sign up, login, and authentication.
Amazon SES can be used to send password-reset emails.
Amazon S3 can be used to store the club documents.
Amazon Bedrock can be used for the AI part of the application and the question-answering system.
The backend can provide the chat API and connect the user's question with the document search and AI response process.
A possible deployment flow would be:
Member ↓ Web Application ↓ Amazon Cognito ↓ Chat API ↓ Document Search ↓ Amazon S3 + Club Documents ↓ Amazon Bedrock ↓ Answer + Source
This is our planned AWS deployment approach. Our current prototype can be demonstrated locally, while these AWS services explain how we could move the application to the cloud.
Challenges
One of the main challenges we considered was making sure that the chatbot does not provide information that is not available in the club documents.
An AI chatbot can sometimes try to answer a question even when it does not have enough information. For our project, we wanted the chatbot to behave differently.
It should use the provided documents and, when the information cannot be found, give the fallback response instead of guessing.
Another important part was making sure that only logged-in members can access the chat page.
What We Learned
Through this project, we learned about building a web application with authentication and an AI-based question-answering feature.
We also learned that an AI application should not only generate answers. It is important to control the information used by the AI and provide the source of the answer.
The project also helped us understand how a local application can be mapped to AWS services such as Amazon Cognito, Amazon S3, Amazon SES, and Amazon Bedrock.
Future Improvements
In the future, we would like to improve the project by adding:
- Chat history for individual members - Thumbs up and thumbs down feedback - Better document search - More club documents - A better mobile-friendly interface - Complete AWS deployment - Easier access to the source documents
Conclusion
Building the Club Member Portal was a good learning experience for our team.
Our main goal was to create a simple platform where Student Builder Group members can log in, ask questions, and find useful information without searching through many documents.
The most important feature of our project is that the chatbot is designed to use the provided club documents as its source and give a fallback response when the required information is not available.
This project also helped us understand how an application can start as a local prototype and later be designed for deployment using AWS services.
We look forward to improving the project further and learning more about building useful applications with AWS and AI.
Any opinions in this article are those of the individual author and may not reflect the opinions of AWS.
Enjoyed reading this content? Let the author know!

Your likes, comments, shares, and saves help creators reach more builders.

Loading recommendations

Loading article