Guardia AI | Devpost

Inspiration

Guardia was born out of the growing crisis of financial scams targeting seniors and vulnerable populations. In Canada alone, seniors lose millions annually to phone and online fraud. These scams are often emotionally and financially devastating, and current tools are reactive rather than preventative. We recognized a gap and a need for a proactive, user-friendly solution that could step in before the damage is done. Our mission with Guardia is to empower caregivers, families, and organizations with tools that detect, prevent, and respond to fraud in real time especially in high-risk communities.

What it does

Guardia is a Chrome extension designed to protect seniors and vulnerable individuals from scams and fraud online. It uses AI-powered scam detection to analyze browser content in real time, flagging suspicious forms with Google Gemini AI . When a threat is detected a pop-up comes on the screen explaining how the page may be a scam. For those who want deeper protection, Guardia offers a premium platform with online training modules and onboarding kits that teach users and families how to recognize and respond to scams. For organizations, Guardia AI provides a partner dashboard and a Care Plan Partnership Program, a revenue-sharing model where caregiver networks and nonprofits can promote Guardia and earn 30–50% per paying user. Partners receive co-branded tools, onboarding support, and scam trend insights, making Guardia a collaborative, scalable solution for digital fraud prevention.

How We Built It

We built Fraud Protector as a full-stack Chrome extension using a modern web development toolchain. The frontend consists of a React application for the extension popup and a content script that runs on every web page, extracting key information such as the page title, meta description, visible text, forms, links, and scripts. This data is then sent via Chrome’s messaging API to a background script, which communicates with our Node.js/Express backend server. The backend is responsible for securely interfacing with Google’s Gemini AI API to analyze the page content for signs of fraud, phishing, or scams, and with Google Cloud Text-to-Speech to generate audio warnings in the user’s preferred language. We implemented robust data parsing and summarization logic to handle large and complex HTML documents, ensuring that only the most relevant information is sent for analysis. The backend returns structured results, including a risk summary, confidence score, and audio alert, which the extension then displays to the user through both visual popups and voice notifications. Throughout the build process, we leveraged technologies such as React, Node.js, Express, Chrome Extension APIs, Google Gemini AI, and Google Cloud TTS, integrating them into a seamless, user-friendly security tool.

Challenges we ran into

One of the biggest challenges we faced was handling and processing large HTML data from modern web pages. Many sites generate complex, dynamic content, resulting in massive HTML payloads that can easily exceed API limits or cause performance bottlenecks. We had to devise strategies to efficiently extract, truncate, and summarize the most relevant parts of each page, such as forms, links, scripts, and visible text,while still providing enough context for the AI to make accurate fraud assessments. Additionally, integrating multiple asynchronous systems (Chrome extension messaging, backend API calls, and third-party AI services) required careful coordination and robust error handling to ensure a smooth user experience.

Accomplishments that we're proud of

We are especially proud of building a fully functional, AI-powered Chrome extension that can analyze any web page in real time and provide both visual and voice alerts to protect users from online scams. Our solution leverages a modern tech stack, including React, Node.js, Express, Google Gemini AI, and Google Cloud Text-to-Speech, to deliver a seamless and accessible user experience. We successfully implemented multi-language support, customizable settings, and robust fraud detection features, all while maintaining performance and usability. Overcoming the technical hurdles of large data handling, cross-origin communication, and browser extension development has been a rewarding achievement for our team.

What we learned

Throughout this project, we gained a deep understanding of how Chrome extensions interact with web pages and external APIs. We learned how to use the Chrome Extension APIs for content scripts, background scripts, and storage, as well as how to manage permissions and messaging between different extension components. On the backend, we became proficient in building RESTful APIs with Node.js and Express, and in securely integrating with Google’s Gemini AI and Cloud Text-to-Speech services. We also learned best practices for handling CORS, managing environment variables, and deploying full-stack applications that bridge browser and server environments.