GOGO VA Chrome Extension:

Inspiration

GoGo VA was born out of the need for a tool that simplifies writing tasks while offering the power of advanced AI models. We wanted to create an AI assistant that works entirely offline, giving users privacy and efficiency, all without the need for an internet connection.

What it does

GoGo VA revolutionizes the writing experience by offering powerful tools for prompt generation, translation, summarization, and more. Whether you're brainstorming ideas, translating text, or summarizing long documents, GoGo VA is designed to help you get your work done faster and more effectively—all from the comfort of your own device.

How we built it

Building a simple Chrome extension using pure vanilla JavaScript without any framework might seem like a good idea at first, but I quickly realized it can become chaotic, especially when it comes to scalability and managing complex projects. The lack of structure makes it difficult to maintain and expand. After investing a significant amount of time in building the core structure, I decided to set up an Angular project. This decision made the development process much smoother, faster, and more organized, allowing me to handle complexity with ease as the project grew. although angular have a steeper learning curve when compared for example react and Vue but I believe it's worth it.

Challenge we ran into

I dedicated significant time to building a text-checker component that seamlessly integrates with text areas as a shadow, offering smart suggestions for grammar corrections, spelling fixes, and more. However, I quickly realized that the AI prompt I was using wasn’t powerful enough to handle this task effectively.

I experimented with a Large Language Model (LLM), which showed promising results compared to the smaller nano models. While the nano models sometimes performed well, they often lacked reliability.

Our goal is to build a ''fully offline' Chrome extension, so instead of integrating the text-checker directly into the app, we made a change in our plan that provides users with feedback on their writing. This approach ensures that users can improve their writing without needing an internet connection.

This experience led me to believe there's a real need for a specialized, tailored nano model (API) to address these kinds of tasks, especially when privacy is a key concern. I’m convinced this could become an essential tool in the market for anyone looking for robust, private, and accurate text analysis.

Accomplishments that we're proud of

We’re incredibly proud of creating a chrome extension that runs 100% offline. The interface is simple and intuitive, making powerful AI accessible to all users.

What we learned

using Angular to build a Chrome extension provided me with a solid foundation for managing complexity, improving code maintainability, and accelerating development. It was a great learning experience that reinforced the power of modern frameworks for building scalable, efficient applications. in addition, expert nano model can preform great when compared to LLM in the same domain of nano model.

What's next for GoGo VA

The extension, powered by the official Gogo VA, currently excels in the area of text. But that's just the beginning! Imagine a future where developers can effortlessly deploy their own nano models, creating a dynamic marketplace. Users will be able to browse and download the perfect model for their needs with just a few clicks. The possibilities are endless, and we're excited to bring this vision to life!.

By integrating the best AI models together, we are working towards creating a fully autonomous AI agent that can not only understand and process human speech but also seamlessly interact with and control a web browser. This ambitious vision combines cutting-edge natural language processing (NLP) models, speech recognition technologies, and browser automation to create an intelligent assistant that can listen to voice commands, interpret them accurately, and take action within the browser.

For example, you could simply speak to the AI agent, instructing it to open a specific website, fill out a form, send an email, or even summarize a page. The AI agent would be able to process these commands, leveraging advanced NLP to understand context and intent, and then control the browser to execute those tasks without any manual input required.

This integration would unlock a new level of convenience and accessibility, allowing users to perform tasks more efficiently and intuitively—whether they're multitasking, managing workflow, or even navigating complex web applications—all with simple voice commands. It's the next step toward making digital interactions smarter, more natural, and hands-free.

Finally, this my vision in the future and currently this is easier to done using LLM but using several nano models that would challenging for sure without your help, invest, plan, etc. would be impossible .

Built With

• angular.js
• chrome
• custom-element
• rxjs
• typescript