Vapi

What it does

The Vapi skill enables AI agents to interact with a voice AI platform for building phone agents and connect them to Getflowetic for monitoring call performance, tracking costs, and analyzing conversation outcomes. It provides access to key metrics related to calls, assistants, and associated costs. The skill normalizes call data into a standard event schema, making it easier to analyze trends and identify areas for improvement in voice agent performance.

When to use it

• Analyzing call volume trends over time.
• Tracking the success rate of Vapi assistant interactions.
• Monitoring per-call costs and identifying opportunities for optimization.
• Comparing the performance of different assistants or workflows.
• Investigating reasons why calls ended (e.g., hangup, error).

Key capabilities

• Access to call data including transcript, status, duration, and cost.
• Normalization of call data into a standard event schema.
• Provides key metrics such as Call Volume, Success Rate, Avg Duration, and Cost per Call.
• Supports dashboard creation with templates like voice-analytics, assistant-performance, and cost-tracker.
• Mapping of legacy field names to canonical Vapi fields (e.g., "duration" maps to "duration_ms").

Example prompts

• “Show me the call volume trend for the last week.”
• “What is the average duration of calls handled by assistant 'SalesBot'?”
• “How many calls resulted in a status of 'success'?"
• "Display a cost tracker showing trends over time."

Tips & gotchas

• The skill requires integration with Getflowetic for full functionality.
• The platform field will always be "vapi" and can be skipped during analysis.
• Be aware of the mapping heuristics; aliases like "duration," "status," and "cost" resolve to specific canonical fields.