Inspiration
In an increasingly digital world, accessibility and inclusivity are paramount. However, older adults often face significant challenges when navigating online platforms and technologies. The digital divide can further isolate them, hindering their ability to access vital information, connect with loved ones, and engage in various online activities. Recognizing this issue, Aly was born out of a desire to bridge this gap and empower older adults to navigate the digital landscape with confidence and ease.
Overview:
Aly was developed with a specific focus on improving accessibility for older people. It embodies the core principles of simplicity, intuitiveness, and minimal user interaction. By understanding the unique needs and preferences of older adults, Aly aims to make browsing and online tasks more accessible and enjoyable, regardless of technological experience.
What It Does
Enhancing User Experience
Aly leverages its intelligence to create clear and concise directions for browser agents. Through a user-friendly interface, older adults can easily communicate their desired outcomes and let Aly take care of the rest. Aly's ability to autonomously discover content, interpret page text, and make educated guesses ensures a seamless browsing experience, reducing the need for excessive user input.
Simplifying Tasks
With Aly, complex tasks that may have seemed daunting for older adults become more approachable. Whether it's managing online accounts, shopping, or researching, Aly breaks down each step into easily achievable actions. Older users can rely on Aly to navigate websites, click buttons, and input text, allowing them to accomplish their goals with confidence and independence.
Promoting Inclusivity
Aly's overarching goal is to foster inclusivity by empowering older adults to actively participate in the digital world. By minimizing the digital divide, Aly opens doors to new opportunities, such as accessing online healthcare resources, engaging in social platforms, or pursuing online learning. It strives to create an environment where older adults feel connected, informed, and supported, regardless of their technological proficiency.
How we built it
Aly was developed by leveraging the power of OpenAI APIs, Electron.js, and Node.js, combining cutting-edge natural language processing capabilities, a versatile cross-platform framework, and a robust backend technology stack.
OpenAI APIs: Aly harnesses the intelligence of OpenAI's language models to understand user instructions and generate clear and concise directions for browser agents. By utilizing OpenAI's advanced natural language processing capabilities, Aly can interpret user inputs, extract relevant information, and generate actionable steps that other browser agents can follow. This allows for seamless interaction and efficient automation of browsing tasks.
Electron.js: Aly's user interface is built using Electron.js, a popular framework for creating desktop applications using web technologies. Electron.js enables the development of cross-platform applications, making Aly accessible to users on various operating systems. It combines the power of Chromium for rendering web pages and Node.js for server-side functionality, providing a robust foundation for Aly's user interface and interaction capabilities.
Node.js: Node.js acts as the backend technology for Aly, handling tasks such as communication with OpenAI APIs, managing browser agents, and coordinating automation processes. Node.js offers a scalable and event-driven environment, allowing Aly to efficiently handle multiple concurrent tasks, interact with the browser, and process data from OpenAI APIs.
The development process involved integrating OpenAI APIs with the Electron.js and Node.js frameworks. This integration enables seamless communication between Aly's user interface, the underlying browser automation logic, and the natural language processing capabilities provided by OpenAI. Aly's frontend, built with HTML, CSS, and JavaScript, interacts with the backend Node.js server, which orchestrates the browser agents' actions based on user instructions and OpenAI-generated directions.
Through the synergistic combination of OpenAI APIs, Electron.js, and Node.js, Aly offers a user-friendly interface, robust automation capabilities, and intelligent decision-making to enhance the browsing experience for older adults and improve accessibility in the digital realm.
Challenges we ran into
The biggest challenge we ran into when building this project was with prompt engineering to configure the GPT-4 model to work best agnostic of the site. Many sites are designed quite differently, and we are very proud of the fact that we were able to configure our app to work on many different platforms.
What we learned
We learned that building a sophisticated browser automation tool like Aly required an iterative and learning-focused approach. The team discovered the importance of continually gathering feedback from tests of the system, analyzing our responses and attempting to understand what exactly GPT-4 was thinking, and using that knowledge to refine Aly's algorithms and prompts. This iterative process helped enhance the accuracy, reliability, and usability of Aly over time.
What's next for aly.so
We want to expand and test the breadth of tasks that Aly can accomplish. We also are exploring ways we can improve the speed of the OpenAI api calls to improve the execution time of a prompt. Furthermore, we would love to see Aly expand to be an entire system's personal assistant rather than just that of a web browser.
Log in or sign up for Devpost to join the conversation.