Innovation rarely follows a linear path. It often begins with a placeholder idea, takes a few messy turns, and transforms into something meaningful. That’s exactly what’s happening with our project, now called AI Policy Pathway.

From Placeholder to Purpose
During this meeting, Luke brought up a great point about the need for questions that lead to answers that the educators are seeking. The team felt that we are at the critical point of being able to deliver that to our target audience.

Designing with Integrity: Grammarly as a Model
Luke suggested that rather than generating a wall of text, the tool’s output should feel more like Grammarly, where each section is not only readable but also traceable back to source material. While this feature will take technical work, it’s where we’re heading.

In the short term, we’re focused on better prompt engineering to get to the final product quicker. Our current system uses one overarching message to generate entire policy documents. But as Luke pointed out, this approach sacrifices quality. The future version will be modular, tailored to the user’s role and context. We’ll also implement a second pass to ensure that all the pieces are logically connected and free of contradictions.

The Moonshot: AI That Cites Its Sources
Luke also suggested that we should build a tool that cites its sources, pulling directly from the documents we’ve reviewed, which will bring real credibility to the tool’s suggestions. The team recognized this will take a lot more technical skills, but they agree with the need.

To get there, we’ll explore Retrieval-Augmented Generation (RAG), a technique that lets the model “read” smaller, digestible chunks of documents instead of being overwhelmed by entire files. This keeps costs down, respects token limits, and improves reference accuracy.

Naming the Project: Finding Our Identity
Because we felt that the project was in good place, we wanted to brand our project. After consulting a few different LLMs to come up with a name that we liked, we landed on a name that captures both purpose and humility: AI Policy Pathway. We wanted to choose a name that embodied this as a  starting point not the final destination. We purchased the domain for it, and we plan to launch the site at some point. 

ISTE Competition Entrance
I let the students know that I would be entering the project into the ISTE Competition, and the need to meet the April 18 competition deadline .

A Culture of Thoughtful Innovation
What struck me most in this meeting was Luke’s emphasis on thoughtful AI, a tool that doesn’t do the thinking for you, but instead invites deeper engagement. In his words, “We want a starting point, and we want everyone who uses this to practice their own critical thinking.”

It’s this mindset that sets our team apart. We’re not building just another AI tool. We’re designing a pathway. An entry point for smarter, more grounded education policy. And with our amazing students leading the way, I have every confidence we’ll get there! 

We invited Skylar Payne, a former student of mine. He is an AI engineer with over a decade of experience, having worked at Google, LinkedIn, and on FDA-cleared AI medical devices, among others. Skylar offered feedback on our current project and the prototype website. Skylar was very complimentary towards our students and their reasoning for the project, and he provided them with great feedback. 

1. Project Overview

• The tool aims to generate AI policy drafts based on user inputs. I worked on simple questions and answers that will generate a fully editable document.
• Due to a lack of existing policy examples from schools, the team pivoted from a scraping approach to a template-based, question-answer mapping method. Yash was able to program it to pull the sections from an individual section in the template document. 

2. Skylar’s Feedback

• Skylar praised the structured input/output method and emphasized dynamic in-context prompting to avoid baked-in bias from static prompts.
• He suggested mapping user inputs to tailored policy text that will give much more intentional output.
• He encouraged building a systematic quality assurance loop: track inputs and outputs, evaluate success as “yes/no,” and grow a set of good exemplars over time.
• He warned against over-relying on rating scales, which we originally thought about creating. He recommended using binary evaluations to reduce subjectivity.

3. Next Steps for the Team

• Reverse-engineer the exemplar policy: For each section, determine what specific question(s) would generate that text.
• Align existing brainstormed categories (e.g., AI literacy, privacy) with clearly defined user-friendly buttons or scale inputs.
• Create a reliable transformation logic that maps multiple user inputs into one output designation (e.g., mild/medium/spicy).
• Evaluate the prompt performance based on how closely outputs match the intended tone/content, refining questions and context accordingly.

4. Collaboration and Planning

• Skylar is open to reviewing code and contributing ongoing technical support.
• The team will create one question per policy section before the next working session.

Moving Forward

• We were pleased with the prototype. 
• We wanted to systematize the logic that links user input → prompt design → accurate, useful policy output.
• All refinements should be aimed at creating a repeatable, user-friendly experience for school leaders that doesn't overwhelm or confuse them.

Luke shared a live demo of the tool, which asks users a series of questions about their school context. It’s clean, intuitive, and user-friendly.

Claire shared how easy it was to use. Sean and Yash suggested features like post-generation editing and chatbot-assisted refinements. Dhruv emphasized intentionality in how we structure questions, breaking down broad categories like "bias" into meaningful, actionable components. 

We discussed not just technical improvements but deeper philosophical design choices. For example, should schools be asked to reveal their names? Would that make administrators uncomfortable? Would an anonymous, template-style output offer more privacy—and in turn, more adoption?

We discussed creating a tiered model.

• Tier 1: A basic draft generated from user inputs.
• Tier 2: Customization options like color-coded recommendations or editing tools.
• Tier 3: A premium, concierge-style experience with a chatbot, school branding, and tailored support.

What Comes Next
The students have now divided into three key task forces:

1. Refine the Policy Template – Reviewing and expanding categories to include AI Literacy, Accessibility, and Transparency.
2. Build the Ideal Output – Creating a “perfect” policy outline to guide prompt engineering and tool design.
3. Research Legal and Ethical Contexts – Mapping state and federal laws, age-based guidelines, and best practices.

We’re also evaluating various large language models (LLMs), such as Claude, Gemini, and ChatGPT, to determine the best fit for our needs.

On Monday, February 24th at 8:00 p.m., the team met to discuss the progress. Because so many of the students have athletic practices and other after-school commitments, the meeting time was set late. In some ways, that forced us to be very efficient. Still, we met until 9:00 p.m. 

1. Organizing the table
Based on Mr. Speiser's recommendations, we chose to group the critical items into three levels from the least restrictive (a.k.a. mild) to the most restrictive (a.k.a.spicy) and somewhere in the middle (a.k.a. medium). We chose to anchor the information on the "Framework for Administrators and Decision-makers on AI Implementation in Schools," a document I created with my colleague Sofia De Jésus. 

2. Setting task deadlines
The group decided to finish the document by Wednesday, February 26th so that the software group could work on creating a prototype. 

3. Next steps
We set another meeting for Tuesday, March 4th so that we can check in about our progress. We also set up a Google Classroom to house all the materials and documents. 

On Thursday, February 13, our students and I had the honor of inviting Leonard Speiser. Our students had an opportunity to "pitch" their ideas and receive guidance. 

As a result of this meeting, we have chosen to focus on five major concepts related to ethical use of AI based on the Framework that Sofia De Jésus and I have worked on as part of the CSTA Equity Fellowship. 

Sometimes I have to pinch myself on how lucky I feel to work in this incredible community where I can actually have our students interact with someone like Leonard who "has founded six technology companies including Bix, which sold to Yahoo! in 2006, and Clover, a payment company that currently processes over $100 billion a year for small businesses that sold to First Data/Fiserv. [He] led product teams at eBay in 2000 and Intuit in 1998" (https://www.aspentechpolicyhub.org/team/leonard-speiser/). He was gracious enough not only to spend time with our students but also offered to come back and provide additional feedback on our project. 

As we continue working on the project, we had an additional member join the team. Dhruv who wrote the application for the Digital Well-Being Challenge joined the crew with his prompt engineering skills. 

Luke and Axel also created the prototype of what they thought the codes should be. 

Finally, the discussion continued with whether a chatbot would be more appropriate or a template made better sense. 

On Friday, January 17, Jamie Nuñez from Common Sense Media visited the students to discuss the possibility of working with our students. Although there were a few other students who are not necessarily a part of the Innovator Project attended the meeting, I felt that the meeting encouraged the core team members to continue working on the project.

During our conversation, Jamie talked about the AI courses that Common Sense Media has produced and published for educators. He spoke of various opportunities that Common Sense Media offers for educators and students on AI literacy development. After the students left, Jaime and I talked about a few different ideas that I had for our students such as a hosting a panel discussion on AI. 

I have no idea what will come of our conversation, but I am looking forward to continuing our partnership with Common Sense Media. 

 To begin, I put out a call to a small group of students who participated in both the MVHS Tech Internship Program and the Principal's Advisory Council.  In addition, I reached out to Leonard Speiser, one of our parents and a co-lead mentor for our Robotics team who agreed to speak to the students in case they had any questions. 

Initially, Axel Adin and Sean Murphy jumped on the opportunity, and I met with them on Friday, January 10th to form the team.

After our initial meeting, I met with 3 additional students. They are:
Myra Jain
Claire Schwarzhoff
Luke Stephenson

On Wednesday, January 15, we hosted our first strategy meeting, and the students decided to create a website, work on creating a data set, and begin working on technical side of things. They also created a shared folder to house all the work such as a resource document with various useful data set (ex. California Department of Education's Data and Statistics page), a spreadsheet to track the project, a document to brainstorm the need, and a group text for rapid communication.

What was interesting to me was that the students didn't want to use Slack even though they are very familiar with the platform as they use it frequently for other groups. They shared that they would prefer it to keep things simpler before moving to a platform like Slack. 
​
Below are a couple of pictures of them at the start of the meeting. The picture on the right features Axel, Sean, Luke, Claire and Myra (from left to right). We also discussed recruiting more team members. In the end, they decided that it was best to keep the team relatively small at the start and recruit others who can provide specific skills as the project grows. 

In Amsterdam, we were asked to created a pitch for our project. Because we had access to Gemini-powered Google Vids that created this video in minutes. 

As a school leader, I like to believe that I understand some of the challenges that school leaders face. Since the release of ChatGPT in November 2022, many educators have been grappling with the impact of Artificial Intelligence (AI), in particular, Large Language Models (LLM) on schools. I have been thinking and speaking about it in various situations, and I began thinking about how I could help my fellow school leaders who were in search of useful, practical, and safe help in the ever changing AI landscape.  

When I apply to be a part of the Google Innovator program, the application required me to share an idea for a possible future project. Naturally, I chose AI as a topic because I wanted to do things that will support my fellow school leaders. 

By the time I arrived at Amsterdam, I had a pretty good idea on the project I wanted to pursue. I wanted to create a tool that will allow any school leader to create a solid draft for their schools based on where their school is located.