Personal Injury Intake Caller: Retell AI Voice Agent with 5-Question Qualification & Callback Logic

Problem: Kimball Law's accident (MVA/PI) form submissions piled up faster than intake staff could dial, so leads went cold or arrived at specialist calls unqualified and unverified.

Solution:

Built 'Alex,' an outbound qualifying agent on Retell AI powered by GPT 4.1 with the Grace voice. The conversation flow verifies caller identity with first/last name matching rules, checks call timing with a callback-preference branch for busy leads, then runs a five-question qualification sequence (accident type, incident recall, injury status, fault, referral source) asked one at a time with empathetic acknowledgment and anti-repetition logic. An Extract Variable node writes answers into structured intake fields (accident_type, accident_date, injury_status, at_fault, referral_source), a Confirm Contact Details node spells email back character-by-character and reads phone numbers in grouped digits before booking the intake call, and dynamic variables ({{lawFirmName}}, {{intakeSpecialistName}}) make the agent reusable across firms.

Result:

Every form lead now receives an immediate qualifying call at $0.305/min with 1.7–2.2s latency, arriving at the intake specialist verified, qualified, and booked.