An end-to-end AI system that analyzes system/application logs to detect anomalies, predict failures, and provide early warnings.
The project integrates machine learning, NLP (TF-IDF), anomaly detection, and time-ahead prediction, deployed via an interactive Streamlit dashboard.
- Log parsing and feature engineering
- Failure prediction using Logistic Regression
- 30-minute ahead failure prediction
- Anomaly detection using Isolation Forest
- NLP-based log message analysis (TF-IDF)
- Auto-generated incident summary
- Interactive Streamlit dashboard
Raw Logs ↓ Log Parsing & Cleaning ↓ Feature Engineering (Severity, Error Counts, TF-IDF) ↓ ML Models ├─ Failure Prediction ├─ Time-Ahead Prediction ├─ Anomaly Detection └─ NLP-Based Prediction ↓ Decision Layer & Incident Summary ↓ Streamlit Dashboard
| Task | Model |
|---|---|
| Failure prediction | Logistic Regression |
| 30-min ahead prediction | Logistic Regression |
| Anomaly detection | Isolation Forest |
| Log text analysis | TF-IDF + Logistic Regression |
- Python
- pandas, scikit-learn
- TF-IDF (NLP)
- Streamlit
- Git & GitHub
pip install -r requirements.txt
streamlit run dashboard/app.py