Classification · Security Analytics
Next-Gen Intrusion Detection
Explore suspicious network activity through ML predictions, a threat dashboard, and a conversational interface for analyst questions.
- Python
- XGBoost
- scikit-learn
- Streamlit
- Plotly
- RAG
What I built
A hackathon prototype connecting behavioral features with an analyst-facing workflow.
- Trained XGBoost and scikit-learn models over behavioral network features.
- Built Streamlit and Plotly views for threat visualization and triage.
- Added a retrieval-based conversational interface for natural-language threat questions.
How it works
- Prepare network behavior features for classification.
- Run the trained model and expose predictions to an analyst dashboard.
- Visualize suspicious activity for triage.
- Retrieve context for conversational threat queries.
Evaluation & evidence
A reproducible dataset split and confusion matrix are not available in the public material. No accuracy figure is presented as a validated benchmark here.
Scope & limitations
A prototype, not a deployed security product. Source and an end-to-end public demo are not currently linked.