Reach

Inspiration

Our project is called Reach.io, and it was created to solve a problem that many students face: finding research opportunities. We noticed that students often spend hours searching through university websites, reading professor biographies, and trying to figure out who to contact. Universities have hundreds of professors, making it overwhelming for students to manually find the right match. Current methods require students to spend hours browsing websites and writing individual emails, which is difficult for people with little to no research experience.

What it does

Reach.io first takes in a short survey of the student about thier intersts then sources research papers and experienced professors mainly using OpenAlex, a library for research papers. Then using OpenAI it automates sending cold emails to those professors saving much work and time.

How we built it

We designed Reach.io as a pipeline where information flows through multiple stages, with each component responsible for a specific part of the process. The frontend handles user interaction, while the backend coordinates the AI agents and manages the flow of information. Using structured JSON responses allowed our agents to communicate reliably and ensured that information could move smoothly throughout the application. Our application is scalable, and redundant.

Challenges we ran into

Throughout the project we were faced with several challenges that tested our problem solving and creativity skills. One of our biggest challenges was integrating our AI agents through the Fast API backend, and Reach API. Each agent produced data in a different format, so we had to standardize the JSON response to ensure that the output of one agent could be correctly passed to the next agent. We learned that building AI systems is not only about creating prompts, but also about managing data flow and reliability. Moreover debugging these issues took signifiant time and effort. Debugging the communication between multiple systems pushed us to think more carefully about how data moves through an application. Finally the development of the user interface to make it simple, and functional but also provide meaningful information and have proper substance was difficult, as it required propper planning and brainstorming.

Accomplishments that we're proud of

Despite these obstacles, we successfully created a working application that connects students with research opportunities, finding and closing the gap between students and professors so that students can gain opportunities. Also the ability for us to work by ourselves and development our skills made this competition not only help us develop our coding skills but create a working invention that solves a real problem faced by students.

What we learned

Throughout the hackathon, we also learned new technologies and concepts while building the project, including AI integration and learn about full-stack development. We improved our technical skills, and demonstrates the value of combining AI with an proper user experience. We learning more about API, such as the use of Fast and React API in the communication of the front end and back end. We also learned about proper workflow through the use of multiple ai agents, learning about endpoints, and how to carry on information from one agent to the other, and eventually back to the front end.

What's next for Reach

We plan to continue improving Reach.io by expanding its features, improving the AI workflow, and making it even easier for students to discover research opportunities and connect with professors. In the future, we plan on gaining credibility from actual universities and organizations to prove ourselves as reliable while, adding additional features that let professors hop on calls and create 1-1 lessons and mentor students directly.