Usage Analytics & Performance Metrics

Subscribing to Sypha's Teams or Enterprise plans unlocks deep insights into your organization's AI operations. Through the Sypha Gateway, you can track real-time activity, manage budgets, and analyze usage patterns across your entire team.

Overview of the Analytics Dashboard

Navigate to the Usage Details area within your management console to view comprehensive reporting. This dashboard provides a centralized view of all interactions processed through the Sypha Gateway.

[!NOTE] This reporting covers all traffic routed via the Sypha Gateway provider. It does not capture activity from other third-party API providers you may have configured directly in the Sypha extension's settings.

Core Summary Metrics

The top of your dashboard highlights five essential metrics:

• Total Investment: The cumulative expense for the designated timeframe.
• Request Volume: The total count of API calls initiated.
• Average Cost per Query: The mean expenditure for each individual request.
• Token Throughput: The total volume of tokens (input and output) processed.
• Collaborator Engagement: The number of unique team members actively utilizing the service.

Flexible Timeframe Filtering

Analyze your data across several standard intervals:

• Weekly View: Activity over the previous 7 days.
• Monthly View: Trends from the last 30 days.
• Annual View: A retrospective of the past year.
• Full History: Every data point since the account's inception.

Data Presentation & Filtering Options

Personal vs. Team-wide View

The "Only my usage" selector allows you to isolate your footprint:

• Active: View only your own interactions and costs.
• Inactive: View the consolidated data for the entire organization (Administrative access required).

Detailed Breakdown Formats

Choose how you want to visualize your organization's activity:

Daily Distribution

See how usage fluctuates over time with columns for:

• Timestamp: The specific date of activity.
• Spending: Financial total for the day.
• Operations: Number of requests handled.
• Volume: Total tokens (hover for input/output details).
• Users: Number of active contributors.

For team views, you can expand any date to see exactly which team members were active and their specific contributions to the totals.

Model & Date Matrix

Break down activity by the specific AI models in use:

• Date: When the activity occurred.
• AI Model: The specific engine utilized (e.g., anthropic/claude-3.5-sonnet, openai/gpt-4o).
• Expenditure: Cost attributed to that specific model.
• Frequency: Request count for the model.
• Tokens: Tokens processed by the model.
• Users: Total headcount utilizing that specific model.

Deep dive into any row to see individual member statistics for that model and date.

Project-Based Categorization

Sypha can also organize usage data by project.

Project names are automatically detected from the origin remote in your project's .git/config. For instance, a repository URL like git@github.com:organization/project-name.git will be identified as project-name.

Alternatively, you can manually define a project name by creating a .sypha/config.json file in your root directory:

{
	"project": {
		"id": "custom-project-name"
	}
}

Interpreting Your Data

Model Tracking

Sypha provides granular tracking for diverse model architectures, including:

• anthropic/claude-3-5-sonnet
• openai/gpt-4o
• google/gemini-pro
• mistral/developerstral

Team Attribution

When analyzing organizational data, you can identify:

• Activity levels for specific team members.
• Email-based user identification.
• Precise cost and token distributions per person.

Financial Oversight

All costs are reported in USD with high precision, enabling you to:

• Detect spending trends and anomalies.
• Identify which models or time periods are most resource-intensive.
• Quantify the financial impact of each team member's AI usage.

Explore More

• Configure team billing and payments
• Manage organizational roles and access