mAIn Office: KPH

The Arizona State University Artificial Intelligence Cloud Innovation Center, powered by Amazon Web Services (AWS), collaborated with Kelvyn Park Junior and Senior High School to tackle a pressing issue in today's diverse educational landscape: new communication tools are required to meet the needs of multilingual families. These families desire to be involved in the school community and want the latest information available, but face barriers due to English not being their native language. To address this challenge, the AI CIC developed "mAIn Office" – a multilingual chatbot named Luisa designed to provide real-time school information and support in multiple languages, bridging the communication gap and fostering increased parent involvement.

Many schools, like Kelvyn Park, face significant communication barriers with non-English speaking family elders, limiting their involvement and engagement. Traditional methods fail to address the dynamic, multilingual needs of these diverse communities, necessitating an innovative, real-time solution."mAIn Office" is an AI-powered chatbot that leverages advanced natural language processing and translation technologies to support English and Spanish speakers. It integrates with existing school databases and dynamic content to offer accurate and up-to-date information, ensuring that parents, students, and staff can easily access the information they need. The chatbot is trained on various school resources such as weekly newsletters, the student and parent handbook, and the school website making it a comprehensive tool for school-specific information. 

The project involves building and finalizing the architectural solution to ensure a robust and scalable infrastructure. Weekly updates and training iterations based on user interactions enabled the team to continuously enhance the chatbot's accuracy and usability. To develop "Luisa," the bilingual conversational chatbot for Kelvyn Park Junior and Senior High School, the project utilized several key AWS services. Amazon Bedrock was employed to establish the knowledge base that powers the chatbot's conversational abilities. AWS Lambda functions were implemented to manage backend processes, including user query handling and data processing. Amazon S3 was used for secure data storage, particularly for storing emails and documents that feed into the chatbot's knowledge base. Additionally, Amazon Simple Email Service (SES) was integrated to automate the ingestion of information from email communications, ensuring that the chatbot's data remains current.

Just like Kelvyn Park Junior and Senior High School in Chicago’s Hermosa neighborhood, the problem of limited parent involvement due to language barriers is evident in many schools. Traditional methods of communication like newsletters, websites, or even single-language chatbots fail to cater to a multilingual audience and often do not keep up with the dynamic information needs of a school. "mAIn Office" addresses these challenges by providing a real-time, inclusive communication tool that adapts to the user's native language and intent, offering targeted, school-specific information.

"In the summer of 2023, I set out to develop a solution to the decades-long parent engagement challenge we faced as a school due to language and cultural barriers. At its outset, this was simply a “basement project” (shoutout to Colleen Schwab for coining the term) where I used generative artificial intelligence tools to help me write a Python script that would serve as a multilingual school chatbot where parents could text or type their questions about the school in either English or Spanish and receive an accurate response in the user’s native language. Although I had a crude (that’s a nice word for it) yet functional script, it lacked the dynamic component I had envisioned to ensure parents had access to the most up-to-date information that, for a school, is ever evolving. 

In Spring 2024, I heard ASU’s Elizabeth Reilly, Executive Director, AI Acceleration, Enterprise Technology panel at the 2024 ASU+GSV Summit where she described the CIC that specialized in solving public sector problems just like mine. After an engaging and supportive conversation with Nate Wilken and John Rome at the ASU booth, they recommended submitting a proposal to CIC. I am so glad I did. Arun Arunachalam and his team, specifically Priyam Bansal, Tom Orr, Colleen Schwab, Aryan Khanna, Shalini Hanagandi, and Lahari Shakthi Arun breathed life into this “basement project” and converted it into the dynamic solution I dreamed it could be. Insatiable in their drive to design and improve the functionality of the solution, this team obliterated all obstacles that arose along the way in making the multilingual school chatbot, Luisa, become a reality. Working with this team has been an inspiration to me as a public educator because of the intersectionality of all of our work as we collaboratively tackle societal issues that will ultimately enhance the learning experiences of our youth."

-Dr. Keith W. Adams, Principal, Kelvyn Park Junior & Senior High School in Chicago Public Schools

While "mAIn Office" is currently being developed for Kelvyn Park Junior and Senior High School, the potential for wider application is significant. Other schools facing similar communication challenges can benefit from deploying a multilingual chatbot tailored to their specific needs. This solution can be scaled and customized for different educational settings, promoting better parent involvement and fostering an inclusive school community.