🤖 Nova Cost Intelligence: AI-Powered AWS FinOps Suite

A sophisticated cloud financial management tool that bridges the gap between raw AWS billing data and actionable business intelligence. This utility leverages the Boto3 SDK to aggregate multi-service cost metrics and utilizes the Amazon Nova Micro LLM via Amazon Bedrock to generate strategic optimization reports.

---

🌟 Key Features

• Global Spend Visibility: Aggregates daily cost data across all AWS services for a comprehensive 30-day financial snapshot, including a calculated Grand Total.
• Deep Resource Auditing: Specifically targets the "Big Three" of cloud waste:
  • Compute (EC2): Identifies unattached and idle EBS volumes.
  • Database (RDS): Monitors instance status and identifies potential idle DBs.
  • Storage (S3): Flags buckets missing Lifecycle Policies to prevent indefinite storage costs.
• AI-Powered Strategy: Uses Amazon Nova Micro to act as a Virtual CFO, providing:
  • Context-aware "Quick Win" suggestions when waste is detected.
  • Advanced "Growth Strategies" (Graviton, Savings Plans) when the account is already optimized.
  • Dynamic Risk Scoring (1–10) for infrastructure hygiene.
• Professional CLI Dashboard: Utilizes the Rich library for a high-end, terminal-based user interface with real-time status spinners and formatted tables.

---

🏗️ Architecture

The project follows a modular Collector-Brain-Orchestrator pattern:

1. Collector (src/collector.py): The data ingestion engine that interfaces with AWS Cost Explorer, EC2, RDS, and S3 APIs.
2. Brain (src/brain.py): The reasoning engine that processes raw JSON data and applies prompt engineering to the Amazon Nova LLM.
3. Orchestrator (main.py): The application entry point that manages data flow, state aggregation, and terminal rendering.

---

🛠️ Tech Stack

Component	Technology
Language	Python 3.9+
Cloud SDK	Boto3 (AWS SDK for Python)
AI Model	Amazon Nova Micro (via Bedrock)
UI Framework	Rich (Terminal formatting)
Environment	python-dotenv (Credential mgmt)

---

🚀 Getting Started

Prerequisites

1. AWS Account: Ensure you have an active AWS account.
2. Model Access: Enable access to Amazon Nova Micro in the Amazon Bedrock console (typically in us-east-1 or us-west-2).
3. IAM Permissions: Your IAM user needs ReadOnly access to Billing, EC2, RDS, and S3, plus InvokeModel permissions for Bedrock.
4. Cost Explorer: Manually "Launch" or "Enable" Cost Explorer in the AWS Billing console (Note: Data can take 24 hours to populate initially).

Installation

1. Clone the repository:

   git clone https://github.com/YOUR_USERNAME/nova-cost-intelligence.git
   cd nova-cost-intelligence

2. Set up a virtual environment:

   python -m venv venv
   source venv/bin/activate  # On Windows: venv\Scripts\activate

3. Install dependencies:

   pip install -r requirements.txt

4. Environment Setup:

   The script automatically uses credentials from aws configure. If you prefer environment variables, create a .env file based on the provided template:

   cp .env.example .env

   Then fill in your credentials:

   AWS_ACCESS_KEY_ID=your_key
   AWS_SECRET_ACCESS_KEY=your_secret
   AWS_DEFAULT_REGION=us-east-1
   

Running the Audit

Execute the main orchestrator to start the scan:

python main.py

---

🛡️ Security & Best Practices

• Zero Hardcoding: No AWS keys are stored in the source code.
• Git Safety: The .env file and venv/ directory are strictly excluded via .gitignore.
• Least Privilege: The application is designed to operate entirely on ReadOnly permissions.

---

📂 Project Structure

nova-cost-intelligence/
├── src/
│   ├── __init__.py       # Package initializer
│   ├── collector.py      # AWS Resource & Cost data logic
│   └── brain.py          # AI Logic & Prompt Engineering
├── .env.example          # Template for local secrets
├── .gitignore            # Security filters for Git
├── main.py               # Application Entry Point & UI
└── requirements.txt      # Project dependencies