Home

Cloud Insight AI

Cloud Insight AI is an AWS cost + log analysis toolkit that can be used as a Python library or paired with the static dashboard UI.

What you get

• Cost analysis with per-service breakdowns and trends
• Log analysis with error/warning summaries and patterns
• Optional dashboard UI that renders JSON reports

Repo highlights

• src/cloud_insight_ai/ — library code
• data/ — sample cost and log data
• examples/ — usage examples
• tests/ — test suite
• index.html, style.css, script.js — dashboard assets

Quick links

• Getting-Started
• Library-API
• Dashboard
• Configuration
• Development

Getting-Started

Getting Started

Requirements

• Python 3.8+

Install (PyPI)

pip install cloud-insight-ai

Quick start (library)

from cloud_insight_ai import CloudAnalyzer

cost_data = [
    {'service': 'EC2', 'date': '2025-01-01', 'cost': 150.25},
    {'service': 'S3', 'date': '2025-01-01', 'cost': 45.50},
]

logs = [
    '2025-01-01 10:00:00 [INFO] Application started',
    '2025-01-01 10:05:30 [ERROR] Connection timeout',
]

analyzer = CloudAnalyzer()
result = analyzer.run(cost_data, logs)

print(f"Total cost: ${result['summary']['total_cost']:.2f}")
print(f"Errors: {result['summary']['error_count']}")

From source

Dependencies are defined in pyproject.toml.

pip install -e .


### Library-API
```markdown
# Library API

## CloudAnalyzer
```python
from cloud_insight_ai import CloudAnalyzer

config = {
    'cost_thresholds': {
        'high_cost_service_percent': 30.0
    },
    'log_thresholds': {
        'max_error_count': 15,
        'max_warning_count': 25
    },
    'error_patterns': [
        {'name': 'Connection Issues', 'keywords': ['connection', 'timeout']}
    ]
}

analyzer = CloudAnalyzer(config=config)
result = analyzer.run(cost_data, logs)

analyze_cost

from cloud_insight_ai import analyze_cost
cost_result = analyze_cost(cost_data)

Input format

[
  {'service': 'EC2', 'date': '2025-01-01', 'cost': 150.25},
  {'service': 'S3', 'date': '2025-01-01', 'cost': 45.50},
]

analyze_logs

from cloud_insight_ai import analyze_logs
log_result = analyze_logs(logs)

Input format

[
  '2025-01-01 10:00:00 [INFO] Application started',
  '2025-01-01 10:05:30 [ERROR] Connection timeout',
]


### Dashboard
```markdown
# Dashboard

The dashboard is a static site that fetches a JSON report and renders charts.

## Files
- `index.html` (inline JS)
- `style.css`
- `script.js` (alternate JS implementation)

## Data source
The UI expects a JSON report at a URL:

- `index.html` uses: `const S3_URL = '.../final_report.json'`
- `script.js` uses: `const S3_REPORT_URL = '.../final_report.json'`

Update the constant to point at your hosted JSON.

## Hosting
You can host the dashboard on any static host (S3 static website, GitHub Pages, etc.).
Ensure the JSON file is publicly readable and CORS-enabled.

## Expected JSON fields (high-level)
The UI reads fields like:
- `timestamp`
- `cost_breakdown[]` (service, cost, percentage, trend, change_percent, recommendation)
- `cost_trend.history[]` (date, cost)
- `log_levels` / `error_details` / `warning_details` / `info_details`
- `ai_insights[]`

Configuration

# Configuration

## Library configuration (CloudAnalyzer)
```python
config = {
  'cost_thresholds': {'high_cost_service_percent': 30.0},
  'log_thresholds': {'max_error_count': 15, 'max_warning_count': 25},
  'error_patterns': [
    {'name': 'Connection Issues', 'keywords': ['connection', 'timeout']}
  ]
}

Environment variables

Used by configuration helpers in src/cloud_insight_ai/config.py:

- DEBUG (true/false)
- ENV (development or production)
- AWS_REGION
- S3_BUCKET

config.yaml (reference)

The repository includes config.yaml showing a full configuration for:

- cost/log thresholds
- log sources
- Comprehend settings
- notifications
- storage settings


### Development
```markdown
# Development

## Install dev dependencies
```bash
pip install -e ".[dev]"

Run tests

pytest

Formatting & typing (optional)

black .
flake8
mypy src