yash6810/Sentilyze
End-to-end stock sentiment analysis & momentum prediction system using NLP (FinBERT) and technical indicators.
What's novel
End-to-end stock sentiment analysis & momentum prediction system using NLP (FinBERT) and technical indicators.
Code Analysis
1 files read · 4 roundsA Streamlit web application that claims to backtest leveraged stock trading strategies combining FinBERT sentiment analysis with technical indicators, but the core implementation files are inaccessible for review.
Strengths
The project has a reasonable modular structure with clear separation of concerns (data ingestion, feature engineering, modeling, backtesting). It uses standard ML libraries and has a polished Streamlit interface. The code organization follows conventional Python patterns.
Weaknesses
Cannot evaluate core implementation quality due to inaccessible source files. The README claims unrealistic returns (+28,000% to +150,000%) that suggest fundamental flaws in backtesting methodology. No visible test coverage. Cannot verify if the 'Leveraged Alpha' strategy actually implements proper risk management.
Score Breakdown
Signal breakdown
Innovation
Craft
Traction
Scope
Evidence
Commits
70
Contributors
2
Files
452
Active weeks
8
Repository
Language
Python
Stars
1
Forks
1
License
Apache-2.0