Curie

Inspiration

Although we are a team of male-identifying students we were surprised by the lack of representation of women in STEM, a concern voiced by several of our peers, friends, and family. Hence, we were motivated in creating a website that would empower women within the STEM field and hopefully inspire future generations. So, we felt that the only way to understand what women want to see on our website is by asking women themselves. Thus, we collected data from our friends and family within the STEM field. These were the three key concepts they wished to see: Representation of women in STEM, Networking opportunities, Promoting STEM education

What it does

Our website includes a home page that highlights a woman of the week in STEM and allows you to connect with them through Linkedin. The site also finds and showcases articles and videos created by female professionals. Users are able to search their preferred STEM topics and even filter the results based on their interests. They can also choose to find or create blog posts about current developments in the STEM field, or any related topics. The website also locates nearby upcoming events related to women in STEM offering a chance for users to gain opportunities to network.

Another prominent feature of our event is letting users create or find blog posts about current developments in the STEM field or related topics. While creating blog posts, our application utilizes machine learning APIs like OpenAI's gpt3 API and Cohere's API to increase user experience for writers and assist them. With gpt3, we suggest an outline idea for the post and give an introduction paragraph using Cohere API. We also recommend related research papers from the Google Scholar database for the writer to read before starting to write the article.

How we built it

We meticulously built our website utilizing React (HTML, CSS, and JavaScript) for the frontend and Django for backend development. Our project was designed and developed using Figma, which allowed for efficient collaboration and streamlined the design process. The following tools were utilized to make this project a reality:

Tech Stack: Frontend: React Major libraries used on the frontend: Redux, bootstrap, firebase Backend: Django Library: Django Rest Framework

AI APIs: OpenAI's latest gpt3 model text-davinci-003 to generate outline ideas for the writer Cohere API to generate a sample introductory paragraph to a blog post given a blog title, audience, and tone of voice.

Other APIs: SerpAPI to suggest related research papers from Google Scholars according to the given title by the writer

Other APIs for data population: Developed 3 additional python scripts to populate our database. We will be using task schedulers like crontab or Amazon CloudWatch Events to regularly update the external events, articles, and Youtube videos we're providing to users.

Challenges we ran into

The first challenge we had run into was deciding on a topic for our project. With the theme being quite open-ended, and the urge of meeting the sponsor requirements, lots of ideas were bounced off one another during an exciting brainstorming session. This was the first of many hurdles to pass. The next challenge was how to implement the Cohere API, as we wanted to compete for the specific sponsorship category. We looked over the available features provided by Cohere and decided on implementing at least one of the features, thus, generating the outline and first paragraph for our blog posts. Speaking of APIs, trying to populate data with the existing API’s deemed difficult as most weren’t providing the data we were looking or their free features were quite limited.

Accomplishments that we're proud of

Our overall accomplishment that we are most proud of is the completion of our project. It wasn't easy forming a group with individuals we'd never worked with before, brainstorming together, and coding together. But the fact that we were able to complete it on a project that empowers women in STEM is truly an honor we all hold proudly.

What we learned

Our group's skills in coding varied drastically. We had one member who has only ever done an introductory python course in the first year of university. He was guided through working on backend development, and graphical design of the website on Figma. All group members were introduced to new APIs such as Cohere's API.

What's next for Curie

When considering the future, we have big plans for our website. We aim to make it more advanced by incorporating new technologies and creative ideas. Some of the features we plan to add include a summarization tool for blog post descriptions and external articles. Furthermore, we hope to develop our own web scraper with Machine Learning (ML) algorithms to provide more accurate external resources for our users. The web scraper would use the ML model to efficiently locate success stories of women in articles, rather than manually inputting keywords. Finally, we intend to use task schedulers like Crontab or Amazon CloudWatch Events to regularly update the external events, articles, and Youtube videos we're providing to users.