# AALM (Adaptive Activity & Load Monitor)

## Overview
AALM is a real-time event monitoring and burst detection system built with FastAPI, SQLite, and a lightweight WebSocket-based frontend dashboard using Chart.js.

It continuously ingests events, computes real-time activity rates, and detects anomalous “bursts” using an adaptive statistical threshold.

The system is designed to simulate and analyze streaming event loads in real time.

## Features
Real-time event ingestion API (/event)
Live event stream via WebSocket (/ws/events)
Adaptive burst detection using rolling statistics
Dynamic thresholding (mean + 2 standard deviations)
Persistent storage using SQLite
Live dashboard visualization
Background simulation generator (fake event stream)
Burst analytics summary API (/summary)

## Architecture
Backend
Built with FastAPI
SQLite database for persistence
Thread-based background processing
Two loops:
Event generator (simulated load)
Analysis engine (rate + burst detection)

## Core Logic
Rate computation: sliding window aggregation (default 60s)
Thresholding: mean + 2σ over historical rates
Burst detection: state machine tracking high-activity intervals

## Frontend
Simple HTML + JS dashboard
Real-time updates via WebSocket
Visualization via Chart.js

## How to Run Locally
pip install -r requirements.txt
uvicorn app.main:app --reload

Open:

http://localhost:8000

## Deployment

Deployed on:

Render

Live:

https://aalm-dj9f.onrender.com

## Database

SQLite database:s
events(timestamp, source, value)
bursts(start, end, peak, total_events)