Aurelia

Connecting women in STEM to industry mentors

Comment

Inspiration

Women entering STEM fields frequently encounter challenges in establishing meaningful connections and finding peers who share comparable experiences due to the male-dominated nature of these fields. Through conversations with various individuals, including friends and family, we have identified the pressing need for a platform that facilitates connections between women already in the industry, enabling them to learn from their successes and gain insights into their career paths. Aurelia emerged as a solution, envisioning itself as a platform that connects women to individuals who have experienced similar challenges and can share their acquired knowledge to inspire and motivate future generations.

What it does

Aurelia is an AI-powered professional networking platform specifically designed for women in STEM fields. The platform offers intelligent profile creation, allowing users to specify professional details such as education, career status, skills, bios, and upload their resumes for comprehensive information. Additionally, Aurelia employs semantic searching, enabling users to find others using natural language queries rather than specific, challenging searches. For instance, a user searching for “people from Virginia Tech” will be presented with all individuals who have attended Virginia Tech. Furthermore, Aurelia incorporates AI-powered profile recommendations, analyzing users’ profiles to match them with the most compatible profiles based on shared interests and potential for collaboration.

Aurelia boasts a range of social features, including a messaging chat system that facilitates communication between users and an Insight feature. This feature is specifically designed for individuals to share impactful experiences throughout their careers. For instance, a user who has read a particularly inspiring career novel can share it along with a link to its source. Additionally, users can share valuable resources, such as dress code options for new hires, which may be challenging to find elsewhere. These features collectively aim to connect women in the workplace and inspire the next generation.

How we built it

To construct Aurelia, we employed React and Vite for the responsive frontend. Chakra UI was utilized for the polished interface, and protected routes were implemented with context-based authentication. For the backend, a Flask-based API system was created with modular route handling for authentication, profile retrieval, and other functionalities. Supabase, a PostgreSQL database, was employed to store data, and Row-Level Security policies were implemented to ensure data security.

For all AI requests, OpenRouters services were utilized to facilitate rapid model switching without significant backend modifications.

Challenges we encountered

The most significant challenge we encountered was implementing the search feature. We desired the search to be natural and automated, while providing results that users anticipate. Consequently, during our initial attempt, we attempted to utilize LLMs for natural language processing on input and automatically populate filters based on the results. However, we encountered numerous issues with synchronization between the AI, database, and frontend. Consequently, we opted to transition to a semantic search system. This system generates embeddings for each profile, enabling comparison with the embeddings of the search query to identify the most pertinent results. We discovered that this approach was more straightforward and yielded consistent results that aligned with our expectations.

Accomplishments we are proud of

We are proud of the development of the entire project. However, we are particularly proud of creating the small AI features that make a significant impact. From the embeddings with semantic search to the utilization of LLMs for profile recommendations, we found these features to be highly intriguing and are proud to incorporate them into the final project.

What we learned

During the development of Aurelia, we gained valuable insights into various aspects. Notably, we learned how to design user-friendly interfaces and effectively utilize AI beyond the realm of generic chatbots. Our objective was to seamlessly integrate AI features, ensuring they remain unobtrusive and enhance the user experience. This process involved extensive research in prompt engineering and optimizing the application’s flow.

Future Prospects for Aurelia

Although Aurelia was initially conceived as a hackathon project, its potential for expansion and enhancement is substantial. Future developments could include the incorporation of more advanced AI systems, enhanced recommendation algorithms, integration with mobile applications, and the provision of more professional preparation resources. The possibilities for Aurelia are virtually limitless.

Built With

Share this project:

Updates