The challenge: Surge events tripling call volume overnight

Crestfield Insurance Group underwrites auto and homeowners policies across the Midwest, with 180,000 active policyholders. Their claims contact centre handled first notice of loss (FNOL) calls with a team of 14 intake specialists.
Each FNOL call took an average of 11 minutes: verifying policy details, capturing incident information, documenting damage descriptions, and assigning the claim to an adjuster. After a major weather event (hail, flooding, ice storms), call volumes could triple within 48 hours. During these surges, hold times exceeded 25 minutes, and Crestfield routinely brought in temporary staff at $35/hour to manage overflow.
The combination of high base volume, unpredictable surges, and the complexity of FNOL data capture made claims intake the single most expensive line item in Crestfield's contact centre budget.

The solution: Every FNOL call handled instantly, from policy verification to adjuster assignment

Thrumble AI handles all FNOL calls as the first point of contact. The AI verifies the caller's policy using their policy number and personal details, then walks them through a structured incident capture flow.
For auto claims: date, time, and location of incident, vehicles involved, description of damage, injuries reported, police report number, and photos requested via SMS link sent during the call.
For home claims: date of discovery, cause of damage, affected areas, emergency mitigation taken, and contractor access details.
Every claim is scored for complexity and assigned to the appropriate adjuster tier in real time. High-severity claims (injuries, structural damage, total losses) are flagged for priority handling.
The AI speaks in clear, empathetic American English, with the pace and tone adjusted for callers who may be distressed.

The results: The numbers speak for themselves

• 3,200+ FNOL calls processed per month, covering both auto and home lines.

• Average intake time: 2 minutes 45 seconds, down from 11 minutes. A 75% reduction.

• 88% of claims fully captured and assigned without any human involvement.

• During a major hail event in Q2, the AI handled a 280% spike in call volume over 72 hours with zero hold times and no temporary staffing.

• Data quality on FNOL submissions improved: 97% of fields fully completed, compared to 78% under manual intake.

• Adjuster assignment time dropped from an average of 4 hours post-call to instant, real-time assignment during the call.

The commercial impact: $540K saved annually. Policyholder retention up 8%.

$540,000 USD in annual savings from reduced intake staffing and elimination of surge temp costs.
Faster claims processing has reduced average claim cycle time by 2.3 days, contributing to a measurable improvement in policyholder retention (up 8% year-over-year in the first 12 months).
Deployed across all lines in 7 weeks, including compliance review for state-specific FNOL requirements.

75% faster intake. 280% surge handled with zero extra staff. $540K saved annually.