{
  "nbformat": 4,
  "nbformat_minor": 5,
  "metadata": {
    "kernelspec": {
      "display_name": "Python 3",
      "language": "python",
      "name": "python3"
    },
    "language_info": {
      "name": "python",
      "version": "3.11.0"
    }
  },
  "cells": [
    {
      "cell_type": "markdown",
      "metadata": {},
      "source": [
        "# Security Behavior Analytics \u2014 Cohort & Peer EDA\n",
        "\n",
        "**Aria AI Security Data Science Team**\n",
        "\n",
        "Exploratory analysis on synthetic enterprise IAM/access behavior events.\n",
        "\n",
        "> Artifacts are pre-computed \u2014 no external downloads or model training required.\n"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {},
      "source": [
        "import json\n",
        "from pathlib import Path\n",
        "\n",
        "ROOT = Path('.')\n",
        "summary = json.loads((ROOT / 'assets/demo/summary.json').read_text())\n",
        "cohorts = json.loads((ROOT / 'assets/demo/cohort_analysis.json').read_text())\n",
        "risk = json.loads((ROOT / 'assets/demo/risk_analysis.json').read_text())\n",
        "print('Events:', summary['total_events'])\n",
        "print('Users:', summary['total_users'])\n",
        "print('Critical risk:', risk['summary']['critical'])\n"
      ],
      "outputs": [],
      "execution_count": null
    },
    {
      "cell_type": "code",
      "metadata": {},
      "source": [
        "import plotly.graph_objects as go\n",
        "\n",
        "dept = cohorts['dimensions']['department']['cohorts'][:10]\n",
        "fig = go.Figure(go.Bar(x=[c['cohort'] for c in dept], y=[c['event_count'] for c in dept]))\n",
        "fig.update_layout(title='Events by Department Cohort', template='plotly_dark')\n",
        "fig.show()\n"
      ],
      "outputs": [],
      "execution_count": null
    }
  ]
}