Vapi Prompt Engineer: Optimize Voice AI Pacing & Configure Call Analysis Metrics

Scope of Work: Fix Number Grouping/Pacing: Resolve issues where the AI groups digits (e.g., saying "nineteen twenty-five" instead of "one, nine, two, five") and speaks IDs too rapidly. Implement phonetic formatting and punctuation-based pacing. Configure Voice Settings: Optimize ElevenLabs settings (Stability, Clarity, and Style Exaggeration) within Vapi to ensure professional, deliberate speech. Set Up Call Summaries: Create a structured data extraction schema to automatically summarize the outcome of every call into clean, actionable fields. Implement Pass/Fail Metrics: Define and set up "Success Evaluation" criteria within Vapi to automatically flag calls as a "Pass" or "Fail" based on whether specific data points were captured. Advanced Vapi Configuration: Fine-tune the Model, Transcriber (Deepgram), and End-of-call/Silence timeouts to handle long hold times and insurance IVRs effectively. Persona Optimization: Ensure the "low-energy, high-efficiency" professional persona remains consistent across all status branches (Approved, Pending, Not Found, Denied). Required Experience: Proven track record with Vapi.ai, Deepgram (Transcription), and ElevenLabs (Voice). Experience setting up Vapi "Functions/Tools" and Analysis (Summaries and Success Evaluation). Deep understanding of GPT-4o/GPT-4o-mini behavior in voice-stream contexts. Ability to format API data variables for individual digit recitation. To Apply: Please provide a brief example of how you have set up custom Success Evaluation metrics or handled Number Pacing in a previous Vapi project. Apply tot his job