Inspiration
Finding relevant research papers is time-consuming, so we built an AI-powered system to enhance discovery using FAISS-based retrieval, NLP embeddings, and AI validation.
What It Does
Our system scrapes research papers, retrieves relevant ones using FAISS, validates their relevance with OpenAI’s model, and organizes them for easy access.
How We Built It
We collected research papers, embedded them with BERT/SciBERT, indexed them using FAISS, validated them with OpenAI, and built a web/API interface for users.
Challenges We Ran Into
Handling large datasets, optimizing retrieval accuracy, fine-tuning AI validation, and balancing performance with scalability were major challenges.
Accomplishments That We're Proud Of
We achieved highly accurate research recommendations, successfully integrated OpenAI for validation, and built a scalable, user-friendly system.
What We Learned
FAISS indexing improves search speed, BERT embeddings enhance retrieval, fine-tuning AI models boosts accuracy, and user feedback refines recommendations.
What's Next
We plan to expand data sources, enhance AI validation, personalize recommendations with reinforcement learning, and deploy as an API and mobile app. 🚀
Built With
- document
- openai
- rag
- research
Log in or sign up for Devpost to join the conversation.