Navi
Hey! Listen! Your AI-powered fairy companion for AWS cost optimization.
Navi is an intelligent conversational interface inspired by the beloved fairy from The Legend of Zelda. She helps you understand and optimize your AWS spending through natural conversation, providing actionable recommendations based on AWS Well-Architected best practices, FinOps principles, and cloud governance guidelines.
Overview
Navi transforms complex AWS cost data into clear, actionable insights through natural conversation. Simply ask questions about your infrastructure, and receive expert recommendations backed by real-time scan data.
Key Capabilities
| Feature | Description |
|---|---|
| Natural Language Queries | Ask questions in plain English about your AWS costs |
| Real-time Analysis | Recommendations based on your latest scan data |
| FinOps Best Practices | Guidance aligned with FinOps Foundation principles |
| AWS Well-Architected | Recommendations follow AWS Cost Optimization Pillar |
| Multi-Account Support | Analyze costs across all connected AWS accounts |
| Conversation Context | Maintains context for follow-up questions |
Getting Started
Prerequisites
Before using Navi, ensure you have:
- At least one AWS account connected to unusd.cloud
- A completed scan with resource findings
- An active subscription (Pro or Enterprise plan)
Accessing Navi
- Log in to your unusd.cloud dashboard
- Click Navi in the navigation menu (marked with "AI" badge)
- Start asking questions or use the suggested prompts
How to Use Navi
Asking Questions
Navi understands natural language questions about your AWS infrastructure. Here are some effective ways to interact:
Cost Analysis Questions
"What was my EC2 cost last month?"
"Which are my most expensive resources?"
"Show me unused resources and their costs"
Optimization Requests
"What are my top 5 cost-saving opportunities?"
"How can I reduce my EBS costs?"
"Suggest Reserved Instances for my workload"
Best Practice Guidance
"How should I tag resources for cost allocation?"
"What's the best strategy for multi-region cost optimization?"
"How can I implement better cloud governance?"
Using Quick Actions
The interface provides quick action buttons for common queries:
- Monthly Summary - Overview of current month's spending
- Waste Report - List of all unused resources
- Optimization Score - Assessment of your cost efficiency
- Cost Trends - Historical analysis and predictions
Prompt Categories
The sidebar organizes prompts into four categories:
- Cost Optimization - Direct cost reduction opportunities
- Resource Analysis - Deep dives into specific services
- FinOps Best Practices - Industry-standard optimization strategies
- Compliance & Governance - Security and policy-related queries
Understanding Responses
Navi provides structured responses with:
1. Summary
A brief overview of key findings, often starting with "Hey! Listen!" for important discoveries.
2. Analysis
Detailed breakdown including:
- Affected resources with IDs
- Current costs and potential savings (referred to as "treasure")
- Percentage improvements possible
3. Recommendations
Prioritized action items with:
- What to do
- Why it matters (cost impact)
- How to implement
- Risk level (low/medium/high)
4. AWS Best Practice References
Navi cites AWS documentation and FinOps principles using blockquotes like this, providing authoritative backing for recommendations.
5. Data Tables
When comparing resources, Navi presents data in clear tables:
| Resource | Current Cost | Potential Savings | Priority |
|---|---|---|---|
| i-abc123 | $150/mo | $75/mo (50%) | High |
| i-def456 | $80/mo | $30/mo (37%) | Medium |
Best Practices for Using Navi
Be Specific
Instead of: "How can I save money?"
Try: "How can I reduce costs on my development environment EC2 instances in eu-west-1?"
Ask Follow-up Questions
Navi maintains conversation context. After initial analysis, drill deeper:
- "What are my unused resources?"
- "Tell me more about the idle EC2 instances"
- "How do I safely terminate them?"
Combine Queries
For comprehensive analysis:
- Start with overview: "Give me a monthly summary"
- Then focus: "Analyze my RDS spending in detail"
- Then act: "What's the safest way to implement rightsizing?"
Validate Before Acting
Navi provides recommendations, but always:
- Review the specific resources identified
- Check for dependencies or business requirements
- Test changes in non-production first
- Follow your organization's change management process
Underlying Technology
AI Model
Navi is powered by GPT-4o mini, providing:
- Deep understanding of AWS services
- FinOps domain expertise
- Context-aware conversations
- Structured, actionable outputs
- Optimized cost/quality balance - 15-20x more cost-effective than GPT-4 Turbo
Data Sources
Navi analyzes 30 days of historical data for comprehensive recommendations:
- Multiple scan history - All scans from the last 30 days
- Trend analysis - Week-over-week and month-over-month patterns
- Management Report insights - Executive-level aggregated data (if available)
- Service aggregations - Findings summarized by service type
- Top cost drivers - Highest-impact items highlighted
Smart Data Aggregation
To provide accurate recommendations without hitting token limits, Navi uses intelligent data summarization:
| Data Type | What's Included |
|---|---|
| Summary | Total spend, savings, asset counts across accounts |
| Trends | Spend direction, savings opportunities, period analysis |
| Service Breakdown | Finding counts, costs, savings per service |
| Top Items | Top 5 highest-impact resources per service |
| Account Overview | Per-account spend and savings metrics |
This approach ensures Navi has comprehensive context while keeping API costs low.
Knowledge Base
Navi incorporates knowledge from:
- AWS Well-Architected Framework - Cost Optimization Pillar
- FinOps Foundation - Best practices and maturity model
- AWS Documentation - Service-specific optimization guides
- Real-world patterns - Common optimization scenarios
Navi's Personality
Navi brings a touch of gaming nostalgia to cloud cost management:
- 🧚 Helpful companion - Always eager to guide you through your cloud journey
- 💎 Treasure hunter - Refers to cost savings as "treasure" and hidden opportunities
- ✨ Encouraging - Celebrates your optimization wins
- 🎯 Focused - Stays on task while keeping things fun
Privacy & Security
Data Handling
- Scan data is processed in real-time
- No data is stored beyond the session
- Conversations are not logged or retained
- API calls are encrypted in transit
Limitations
Navi:
- Provides recommendations, not automatic actions
- Cannot modify your AWS resources
- Relies on scan data accuracy
- Should be used as guidance, not absolute authority
Troubleshooting
"No recent scans found"
Run a scan first to populate data for analysis. Navi needs scan results to provide personalized recommendations.
Slow responses
Complex queries with large datasets may take longer. Navi is analyzing all your findings to provide comprehensive answers.
Generic responses
If responses seem too general:
- Be more specific in your question
- Filter to specific accounts or regions
- Reference specific services or resource types
FAQ
Q: Which subscription plans include Navi?
A: Navi is available on Pro and Enterprise plans.
Q: How current is the data Navi uses?
A: Navi uses your most recent scan data. Run scans regularly for up-to-date recommendations.
Q: Can Navi make changes to my AWS account?
A: No. Navi is read-only and provides recommendations. You implement changes through the AWS console or CLI.
Q: Is my data shared with OpenAI?
A: Data is processed through OpenAI's API but is not used for model training. See our Security page for details.
Q: Why is it called Navi?
A: Navi is inspired by the helpful fairy companion from The Legend of Zelda series, known for guiding players through their adventures with helpful hints. Just like the original Navi, our Navi guides you through your AWS cost optimization journey!
Related Features
- Management Report - Executive-level cost intelligence reports
- Drift Detection - Automated spending anomaly alerts
- Supported Services - Full list of analyzed AWS services