StemPath

• 

  Home Page

• 

  Results

Inspiration

I wanted to make something that helps students like me figure out which STEM career fits them best. Many people are interested in STEM but don’t know where to start or what field matches their interests. That inspired me to build STEMPath, an AI-powered guide that helps users explore their ideal STEM career paths.

What it does

STEMPath takes in a user’s interests, skills, and goals, then uses AI to find the top three STEM careers that best match their input. It explains what each career is about and gives students ideas for what to study or explore next.

How we built it

I built STEMPath using Python and Streamlit for the web interface. The project uses the SentenceTransformer model (all-MiniLM-L6-v2) to compare user input with a dataset of real STEM careers. It uses Pandas for data processing and PyTorch to handle embeddings. The app is deployed online using Streamlit Cloud, and the full source code is available on GitHub.

Challenges we ran into

Some challenges were learning how to cache embeddings properly so the app runs faster and doesn’t reprocess data every time. I also had to make sure the AI model gave good and relevant career results, which took testing and adjusting. Another challenge was deploying the app and making sure everything worked smoothly online.

Accomplishments that we're proud of

I’m proud that STEMPath actually works and gives personalized results that make sense. It was my first time combining AI with a web app and deploying it for others to try. I also made sure the interface is simple and easy for anyone to use, even if they don’t know much about technology.

What we learned

I learned how to use AI models for text similarity, build an interactive web app with Streamlit, and deploy a working project to the web. I also learned how to clean data, optimize performance, and make the app more user friendly.

What's next for STEMPath

Next, I want to expand the career dataset, add more interactive questions, and include personalized learning paths for each suggested career. I also want to let users save their results and explore college programs or courses related to their top career matches.

Built With

• all-minilm-l6-v2)
• cloud
• github
• pandas
• python
• pytorch
• sentencetransformer
• streamlit