🤖 AI-Powered Incident Response Bot

An automated SRE (Site Reliability Engineering) Assistant that transforms passive log storage into an active incident response system. This bot monitors AWS CloudWatch, scrubs sensitive data for privacy, and leverages Amazon Nova Micro via AWS Bedrock to deliver actionable fixes to Slack in under 60 seconds.

---

🚀 Key Features

Feature	Description
Real-time Log Polling	Continuously monitors CloudWatch Log Groups for ERROR, Exception, or Fail patterns
Privacy-First Scrubber	Uses Regex to mask PII (IP addresses & Emails) before data leaves your environment
AI Root-Cause Analysis	Leverages Amazon Nova Micro to generate human-readable explanations and 3-step remediation plans
Cross-Region Integration	Fetches logs from any AWS region (e.g., ap-south-1) while utilizing AI models in us-east-1
Slack Automation	Delivers instant alerts to your team's channel, significantly reducing MTTR (Mean Time To Repair)

---

📂 Project Structure

incident-response-bot/
├── main.py            # The "Heart"          — Orchestrates the main execution loop
├── analyzer.py        # The "Brain"          — Handles AWS CloudWatch & Bedrock logic
├── scrubber.py        # The "Shield"         — Privacy logic to mask sensitive data
├── config.py          # The "Nervous System" — Environment variable management
├── trigger.py         # The "Spark"          — Test script to simulate cloud incidents
├── .env               # Secrets (AWS Regions, Slack Webhooks — Git Ignored)
├── requirements.txt   # Python dependencies
└── Dockerfile         # Containerization blueprints

---

🛠️ Setup & Installation

1. Prerequisites

Before getting started, ensure you have the following:

• Python 3.10+ installed on Ubuntu
• AWS CLI configured with the appropriate permissions:
  • CloudWatchLogsReadOnlyAccess
  • AmazonBedrockFullAccess
• Amazon Bedrock access enabled for the Nova Micro model in us-east-1
• A Slack Incoming Webhook URL

---

2. Virtual Environment Setup

# Create and activate environment
python3 -m venv venv
source venv/bin/activate

# Install dependencies
pip install -r requirements.txt

---

3. Configuration (.env)

Create a .env file in the root directory:

AWS_REGION=ap-south-1
LOG_GROUP=/aws/lambda/production-logs
SLACK_URL=https://hooks.slack.com/services/YOUR/WEBHOOK/URL

---

🚦 Usage

Start the Monitoring Bot

python3 main.py

The bot will begin polling CloudWatch every 60 seconds. You should see:

🚀 Monitoring /aws/lambda/production-logs for incidents...

Simulate an Incident

In a separate terminal, run the trigger script to inject a fake error:

python3 trigger.py

---