Inspiration
Many students struggle to pick the right hackathon project due to limited time, unclear problem understanding, and lack of AI/ML knowledge. I wanted to create a tool that guides students in real-time, helping them choose innovative and feasible projects quickly.
What it does
Takes a problem statement as input.
Generates practical AI/ML project ideas.
Suggests recommended technologies and tools.
Provides guidance for building impactful solutions.
How we built it
Text Preprocessing: Used NLTK for tokenization, stopword removal, and normalization of input statements.
Project Generation: Applied ML algorithms to analyze keywords and semantic similarity against a curated project database.
Web Interface: Built using Streamlit for an interactive, user-friendly experience.
Deployment: Pushed to GitHub and deployed globally on Streamlit Cloud for easy public access.
Challenges we ran into
Semantic Analysis: Matching input problem statements to relevant projects was tricky initially; solved using keyword extraction + cosine similarity.
Deployment Issues: Setting up dependencies and virtual environments for Streamlit was challenging.
Project Diversity: Ensuring output project ideas were unique and practical required creating a diverse database.
Accomplishments that we're proud of
Built a fully functional AI/ML project suggestion tool from scratch.
Created an interactive web app accessible globally.
Solved real-world student pain points in hackathons, making project selection smarter and faster.
What we learned
Practical experience in NLP and text preprocessing
Using ML algorithms for semantic analysis
Building and deploying a Python web application with Streamlit
Managing version control with Git & GitHub
What's next for AI Hackathon Helper
Integrate large language models (LLMs) for smarter project suggestions
Personalize suggestions based on student skill level
Add project templates or starter code for instant project implementation
Expand the system to cover more domains beyond AI/ML
Built With
- cloud
- cosine-similarity
- for
- github
- jupyter
- ml/nlp
- notebook
- numpy-ai-api:-bytez-api-(for-gpt-powered-project-generation)-version-control-/-deployment:-git
- optional
- programming-language:-python-web-framework-/-interface:-streamlit-ai-/-ml:-natural-language-processing-(nlp)
- scikit-learn-libraries:-pandas
- streamlit
- testing
- tf-idf
- tools:
Log in or sign up for Devpost to join the conversation.