ForMai

Inspiration

As a team, we noticed a common frustration while filling out job applications, admission forms, and online surveys. Many form questions are written in an unclear or ambiguous way, making it difficult to understand what type of response is actually expected. Most existing tools focus on automatically filling forms, which can be risky, inaccurate, and sometimes unsafe. Instead of automating form filling, we wanted to build something that helps users understand forms better. This motivation led us to create ForMai.

What it Does

ForMai is designed as an AI-powered form understanding assistant. The core idea is to help users interpret form questions and understand what the form owner is looking for. In its intended form, ForMai works as a Chrome extension that stays inactive on normal web pages and activates only when a form is detected. Users can select any form question to receive a clear explanation of its meaning, expectations, and how to approach answering it. Users can also upload or capture a screenshot of a form to receive high-level guidance on how to handle the form. ForMai does not fill forms and does not store any user data.

Due to the limited hackathon timeline, the current submission focuses on a working AI-powered prototype that demonstrates the core functionality using the Gemini 3 API, while the Chrome extension is the planned deployment layer.

How We Built It

We focused on building and validating the core AI logic using the Gemini 3 API. The prototype uses Gemini 3’s reasoning capabilities for explaining form questions and its multimodal capabilities for analyzing form screenshots. The logic is designed so that AI calls happen only when the user requests help, keeping the system efficient and privacy-friendly. The Chrome extension architecture was designed conceptually, but full implementation was scoped out due to time constraints.

Challenges We Ran Into

One of the main challenges was balancing ambition with the short hackathon timeline. Building a fully polished Chrome extension while also integrating and testing multimodal AI reasoning was not feasible within the available time. Another challenge was crafting prompts that guide users without generating direct answers. We addressed these challenges by prioritizing a functional and testable prototype that clearly demonstrates the intended user experience and Gemini 3 integration.

What We Learned

Through this project, we learned the importance of using AI to support user understanding rather than replacing user decision-making. We gained hands-on experience with Gemini 3’s reasoning and multimodal capabilities and learned how to design prompts that emphasize guidance over automation. We also deepened our understanding of how lightweight and privacy-conscious browser extensions should be designed.

What’s Next for ForMai

In the future, ForMai will be fully implemented as a Chrome extension with improved form detection accuracy and broader form support. We also plan to add accessibility features, multilingual support, and more context-aware explanations. As the project evolves, maintaining user privacy and safe AI usage will remain our top priority.

Built With

• api
• chrome-extension-api
• css
• gemini
• github
• google-gemini-3-api
• html
• javascript
• screenshot-capture-api