EXL: AI Data Analytics and Document Processing

8 min read

On This Page

EXL is a global analytics and digital solutions company specializing in AI-powered data transformation and intelligent document processing across insurance, healthcare, and financial services.

EXL

$2.09B2025 full-year revenue
57%Revenue from data and AI-led services
2,000+AI-powered workflows in production
250+Pre-built agents and accelerators

Overview

EXL closed 2025 with $2.09 billion in full-year revenue, up 13.6% year-over-year, with data and AI-led services accounting for 57% of total revenue while growing at 21% in Q4 alone. The company has posted 20 consecutive quarters of growth through 2025. Headcount grew 9.8% against revenue growth of 13.6%, a gap that signals improving revenue per employee as AI-led delivery scales. EXL won 65 new clients across full-year 2025 and 21 in Q4 alone, with Q4 dollar volume of wins more than double any other quarter in the year.

That trajectory reflects a structural shift that began years earlier. EXL has transformed from a traditional business process outsourcing provider into an AI-native data and analytics company. The October 2025 launch of EXLdata.ai brought an agentic AI-native data suite developed with Databricks that formalized that positioning. By March 2026, EXLdata.ai supported 2,000+ AI-powered workflows in production for 800+ global clients.

Three major partnerships formalized in early 2026 define EXL's infrastructure strategy. A strategic services collaboration with Google Cloud announced April 2, 2026, targets financial services, healthcare, utilities, and retail, backed by 800+ AI-technical specialists trained on Google Cloud technology. An NVIDIA integration announced March 16, 2026 embeds NeMo Framework, NeMo Agent Toolkit, and NIM Agent Blueprints into EXLerate.ai, delivering 30-50% faster time-to-model via NVIDIA CUDA-X GPU-accelerated processing. EXL was also named AWS's 2025 AI/ML Market Disruptor of the Year and became a globally managed Google Cloud Strategic Services partner.

As Vishal Chhibbar, EXL's Executive Vice President and Chief Growth Officer, stated at the Google Cloud announcement: "This collaboration enables us to scale innovation faster, deepen our Google Cloud capabilities and deliver measurable business impact for our clients worldwide."

For 2026, EXL projects $2.275 billion to $2.315 billion in revenue (9-11% growth, constant currency) and adjusted diluted EPS of $2.14 to $2.19. CEO Rohit Kapoor has noted that talent scarcity, not demand, is the binding constraint on growth: "Our data and AI-led business today is constrained for growth due to talent." A $500 million stock repurchase program was authorized February 28, 2026. EXL expects to be active in M&A, with priorities including data readiness for AI, engineering skill sets, new capabilities, and geographic diversification.

How EXL processes documents

EXL does not position its document processing capabilities as a standalone IDP product competing directly with point solutions. Instead, data extraction and document intelligence are embedded within EXLdata.ai as a data foundation layer that feeds downstream agentic workflows. Buyers should expect a platform sale, with pricing, implementation scope, and outcomes tied to the broader data and AI engagement. No published pricing tiers or model strings for EXLdata.ai or EXLerate.ai were disclosed in any source reviewed. More than 75% of EXL revenue is recurring or annuity-like, indicating the company is transitioning from project-based services toward platform-based engagement models.

The document processing stack is anchored by patents granted between December 2023 and January 2026. EXL disclosed 10 new U.S. patents in February 2026, covering multimodal document extraction, knowledge graph creation, and domain-specific LLM fine-tuning. These patents now underpin an explicit IP pricing component in enterprise contracts, a commercial model shift disclosed publicly for the first time in the Q4 2025 earnings call.

The five core document processing patents are:

  • XTRAKTO.AI™ (U.S. Patent No. 12,260,342, March 2025): multimodal table extraction and semantic search for natural-language queries within unstructured documents.
  • EXL Generic NER™ (U.S. Patent No. 12,253,832, July 2025): context-based named entity recognition within large database files.
  • Knowledge graph creation (U.S. Patent No. 12,481,215, November 2025): dynamic content segmentation, concept and relationship extraction, graph-based embeddings, and real-time updates as source data changes.
  • Contextual text recognition via prompt-guided knowledge distillation (U.S. Patent No. 12,033,408, July 2024).
  • Complex enterprise data structuring via machine learning (U.S. Patent No. 11,842,286, December 2023).

A sixth patent, U.S. Patent No. 12,399,924 (August 2025), covers EXL Insurance LLM™, described as the first industry-specific LLM for bodily injury claims adjudication. Built on more than 10 years of domain-specific labeled data and fine-tuned in partnership with NVIDIA, EXL claims 20-30% higher accuracy in claims and underwriting tasks than "leading general-purpose models." No named competitor or named benchmark was cited, so treat this as an unvalidated vendor claim.

The broader platform combines EXLdata.ai's agentic architecture with multi-agent systems for data discovery, migration, and governance. EXLdata.ai supports language-agnostic processing and 80+ modular agents. The March 2026 expansion of EXLerate.ai added Agent Studio for no-code autonomous agent building and an AI Governance Hub covering 40+ enterprise-grade models with built-in guardrails. Code Harbor™ reduces SAS-to-Databricks migration time by 35% (50% faster in a documented insurance deployment). Smart Agent Assist™ cuts contact center handling time by 12-15%. XTRAKTO.AI™ integrates with LifePRO™ for insurance policy administration and EXLerate.AI for workflow simulation and architecture validation.

Clients report 40% reduction in development costs and up to 50% reduction in time from prototype to production using EXLerate.ai, per EXL's March 2026 NVIDIA partnership announcement. These figures are self-reported and carry no independent verification.

Use cases

Insurance operations

Insurance is EXL's largest revenue segment at $185.8 million in Q4 2025, up 7.2% year-over-year. EXL achieved Leader status in five quadrants of the ISG Provider Lens Insurance Services 2025 report, including GenAI and Agentic AI categories. The company partnered with the Insurance Council of Australia to develop a national AI-powered fraud detection platform launching in early 2026, in collaboration with Shift and NVIDIA.

EXL's largest Q4 2025 win involves a new North American insurance carrier with a distinctive pricing structure. The engagement covers EXLdata.ai for data infrastructure, agentic AI embedded in operational workflows, CX transformation, and back-office re-engineering. Critically, EXL is charging explicitly for IP as a pricing component above time-and-material rates. As Vivek Jetley, President of Insurance, Healthcare and Life Sciences at EXL, stated on the Q4 2025 earnings call: "EXL is able to start charging for our IP. We've built in a component here which is an explicit pricing for the capabilities that we're bringing in and deploying." If replicated across the pipeline, this would structurally improve gross margins over time, moving EXL from a services margin model toward a software-adjacent one.

EXLdecision.ai, launched March 11, 2026, accelerates model development by 30-50% and is directly applicable to underwriting and claims decisioning workflows. EXL ClaimsAssist.ai, launched the same date, targets insurance claims automation end-to-end.

Healthcare analytics

Healthcare was EXL's fastest-growing segment in both Q4 2025 and full-year 2025, reaching $142.2 million in Q4, up 25.9% year-over-year. EXL was recognized as a Leader by Everest Group in Healthcare Data, Analytics and AI Services. Its modular healthcare platform integrates diverse datasets across population health, care coordination, and financial analytics.

Two documented deployments illustrate the document processing capability in this vertical. A Fortune 500 healthcare client automated benefit management workflows and claims processing using EXL's Google Cloud-powered AI solutions, reducing manual effort and generating cost savings (amounts not disclosed). A large healthcare payer built a centralized, governed contract repository spanning structured and unstructured data while aligning contract terms with claims adjudication. CEO Kapoor described this use case as "broadly applicable well beyond healthcare, anywhere where contracts and policies drive downstream operational decisions." EXL also expanded its engagement with a top-five U.S. healthcare payer, deploying an AI-powered payment integrity solution to re-engineer clinical auditing processes.

Banking and financial services

Banking, Capital Markets and Diversified generated $122.6 million in Q4 2025, up 10.8% year-over-year. A consumer lending fintech used EXLdata.ai to migrate from legacy on-premises infrastructure to cloud-native in four months, one of two production deployments disclosed at the platform's launch. EXL's knowledge graph and NER patents are directly applicable to financial document workflows involving contracts, regulatory filings, and loan documentation. The April 2026 Google Cloud partnership explicitly targets financial services as a priority vertical for AI-powered document processing and workflow automation.

Legacy system modernization

EXL's Code Harbor solution achieved 50% faster SAS-to-Databricks migration for a global insurance provider, supporting multiple legacy platforms including BTEQ, HQL, and PL/SQL systems. The platform modernization capability positions EXL for enterprise accounts where document processing is blocked by underlying data infrastructure debt, a common entry point for the broader EXLdata.ai platform sale. EXL also deployed agentic AI for Sonos's IT service management workflows in partnership with AWS, demonstrating applicability outside regulated industries.

Technical specifications

Component Details
Platform architecture Multi-agent AI-native with cross-LLM optimization; EXLdata.ai and EXLerate.ai
Document processing XTRAKTO.AI™ (multimodal table extraction, semantic search); EXL Generic NER™ (context-based named entity recognition)
Knowledge layer Knowledge graph creation with dynamic content segmentation and real-time updates (U.S. Patent No. 12,481,215)
Domain LLM EXL Insurance LLM™ for bodily injury claims adjudication; fine-tuned with NVIDIA on 10+ years labeled data (vendor claim: 20-30% accuracy improvement vs. general-purpose models; no named benchmark)
Agentic AI 250+ pre-built agents and accelerators; Agent Studio (no-code agent building); AI Governance Hub (40+ enterprise-grade models with guardrails)
Decision intelligence EXLdecision.ai: 30-50% faster model development; EXL ClaimsAssist.ai for insurance claims automation
Cloud infrastructure AWS, Databricks, Google Cloud (globally managed Strategic Services partner), Microsoft
NVIDIA integration NeMo Framework, NeMo Agent Toolkit, NIM Agent Blueprints; CUDA-X GPU-accelerated processing; 30-50% faster time-to-model
Partner marketplace 16 EXL solutions listed on AWS, Microsoft, Databricks, and Genesys marketplaces
Data processing Structured and unstructured data across clinical records, diagnostic images, insurance claims, contracts
Security End-to-end encryption, role-based access controls, security-first governance
Integration API-driven, modular open architecture; Genesys CX integration
Deployment Hybrid cloud, on-premises, SaaS
Pricing No published tiers; enterprise platform sale with explicit IP pricing component above time-and-material rates (disclosed Q4 2025)

Resources

Company information

Headquarters: New York, United States Global presence: 13 offices, 63,000+ employees across six continents Founded: 1999 Stock: NASDAQ: EXLS Phone: Contact through website Email: Contact through website Web: https://www.exlservice.com

Investor Strategy Update: May 13, 2026, New York