EXL Service

EXL Service derived approximately 55% of its revenue from insurance operations management — processing claims, managing policy administration, and handling underwriting support for major US and UK insurers. By FY2019, these operations relied on approximately 20,000 processors handling high-volume, repetitive tasks: FNOL intake, claims triage, medical record review, and subrogation recovery. Labor cost represented ~65% of delivery expense. The insurance BPO market was becoming commoditized, with Indian competitors pricing aggressively. EXL's insurance operations gross margin was approximately 32%, below its analytics segment (42%) and under pressure from client rate renegotiations averaging 2-3% annual reductions.

EXL launched its "Digital Intelligence" program focused on its insurance vertical:

• Claims triage automation: Deployed NLP and computer vision models to automatically categorize and route incoming claims. The system analyzed claim descriptions, attached images (vehicle damage, property damage), and structured data to assign severity scores and route to the appropriate handler. This eliminated manual triage for approximately 65% of new claims by FY2022.
• Medical record extraction: Built AI models to extract relevant medical information from unstructured clinical records for workers' compensation and disability claims. The system processed approximately 500,000 medical records annually by FY2022, reducing review time from an average of 45 minutes to 12 minutes per record.
• Subrogation identification: Deployed ML models to identify subrogation recovery opportunities that human processors frequently missed. The system analyzed claim data patterns across millions of historical claims to flag potential recovery cases, increasing subrogation identification rates by approximately 30%.
• Straight-through processing: Achieved end-to-end automation (no human touch) for approximately 35% of simple, low-complexity claims (auto glass, minor property damage under $5,000) by FY2022, with auto-adjudication based on policy terms and damage assessment.

• Delivery cost reduction: Cost per claim processed declined approximately 22% from FY2019 to FY2022, from an average of $18.50 to $14.40 per claim across the portfolio.
• Insurance operations margin improvement: Gross margin in the insurance operations segment improved from approximately 32% to 35.5% over the period, a 350 basis point gain.
• Throughput increase: Claims processing throughput increased approximately 28% without corresponding headcount increase — the same workforce processed significantly more volume.
• Quality improvement: Claims processing accuracy improved from 96.2% to 98.5%, reducing costly rework and client penalties.
• Revenue impact: Rather than reducing headcount, EXL redeployed freed capacity to take on additional client volume, growing insurance operations revenue from approximately $780M to $880M (FY2019-2022) with minimal headcount addition.
• Timeframe: FY2019-FY2022 (3-year program).

• Deep insurance domain expertise built over 15+ years of claims processing gave EXL the data and process knowledge to build effective AI models
• Proprietary training data from millions of processed claims enabled high model accuracy from launch
• Analytics segment (EXL Analytics) provided in-house data science talent for model development
• Client partnership model: key insurance clients co-invested in automation development, sharing data access and absorbing some implementation risk