EXL to Acquire iMerit in Up to $310 Million Deal to Expand Enterprise AI and Foundation Model Capabilities

EXL has announced a definitive agreement to acquire AI data specialist iMerit in a transaction valued at up to $310 million, marking one of the company’s most significant investments in artificial intelligence to date. The acquisition strengthens EXL’s position in enterprise AI by adding advanced capabilities in foundation model training, evaluation, reinforcement learning, and multimodal data operations while extending its reach into rapidly growing sectors such as robotics, autonomous systems, and physical AI. The transaction is expected to close during the third quarter of 2026, subject to customary regulatory approvals and closing conditions.

Combining Enterprise AI with Foundation Model Expertise

Founded in 1999, EXL has evolved from a business process outsourcing company into a data and AI-focused enterprise technology provider serving industries including insurance, healthcare, banking, retail, communications, and energy. Today, the company employs more than 67,000 people globally and has increasingly focused its strategy on helping organizations operationalize AI across core business functions rather than deploying isolated proof-of-concept projects.

That strategy has centered around EXL’s growing portfolio of AI platforms, including EXLerate.ai for agentic AI workflow orchestration, EXLdata.ai for enterprise data management, and EXLdecision.ai for decision intelligence. These platforms are designed to integrate AI models, enterprise data, and human expertise into production business workflows while emphasizing governance and industry-specific deployments.

The addition of iMerit fills an increasingly important gap in that ecosystem: the creation and refinement of the AI models themselves.

What iMerit Brings to the Deal

Unlike companies focused primarily on deploying AI applications, iMerit specializes in producing the high-quality training data and expert human evaluation required to build and improve modern foundation models.

The company works with frontier AI developers and enterprise customers to train, fine-tune, evaluate, and align large language models and multimodal AI systems. Its expertise spans text, images, video, audio, LiDAR, and other complex data types used in applications ranging from generative AI to autonomous vehicles and robotics.

A key asset in the acquisition is Ango, iMerit’s proprietary collaboration platform. Ango enables distributed teams of subject matter experts to perform sophisticated AI evaluation tasks including reasoning assessment, multimodal annotation, red teaming, model alignment, and reinforcement learning from human feedback. The platform is designed to support increasingly complex evaluation workflows required by today’s large foundation models.

Equally significant is iMerit’s Scholars network, which consists of physicians, scientists, engineers, linguists, and other domain specialists who provide expert human feedback during model development. Rather than relying solely on large-scale crowd work, iMerit emphasizes expert knowledge for applications where factual accuracy and domain-specific reasoning are critical, particularly in healthcare, scientific research, and other regulated industries.

Strengthening EXL’s End-to-End AI Stack

Following the acquisition, EXL plans to integrate Ango directly into its existing AI platforms.

The combined platform would connect enterprise data preparation, AI orchestration, model evaluation, reinforcement learning, and business workflow execution into a single environment. Rather than treating model development and enterprise deployment as separate activities, the combined technology stack aims to allow organizations to continuously improve AI models using expert human feedback while deploying those models within production systems.

The acquisition also gives EXL direct relationships with foundation model developers, providing greater visibility into how emerging AI models are trained, evaluated, and adapted for enterprise use. According to the company, this should strengthen its ability to help customers develop domain-specific language models tailored to proprietary datasets and specialized business processes.

Expanding Beyond Traditional Enterprise Markets

Although EXL has historically concentrated on industries such as insurance, banking, and healthcare, iMerit’s customer base expands its exposure into several rapidly growing AI sectors.

The company has built expertise supporting AI systems used in autonomous mobility, robotics, industrial automation, computer vision, and physical AI. These applications require highly specialized multimodal datasets that combine camera imagery, LiDAR, sensor fusion, video, and environmental understanding to train models capable of operating safely in real-world environments.

As AI development increasingly extends beyond text generation into machines that perceive and interact with the physical world, the demand for accurately labeled and expertly validated training data continues to grow. iMerit’s experience across these modalities broadens the range of industries EXL can address with its AI offerings.

A Broader Shift Toward AI Infrastructure

The acquisition also reflects how the enterprise AI market is evolving.

Much of the industry’s attention during the past several years has centered on increasingly powerful foundation models. However, organizations are discovering that deploying AI reliably often depends just as much on the surrounding infrastructure, including high-quality training data, continuous evaluation, governance, reinforcement learning, and domain-specific expertise.

For regulated industries such as healthcare, insurance, financial services, and life sciences, these capabilities are becoming increasingly important as organizations seek AI systems that can be adapted to specialized workflows while meeting requirements for accuracy, transparency, and oversight.

By bringing together enterprise workflow automation with model development expertise, EXL is positioning itself to participate across a larger portion of the AI lifecycle—from preparing and refining models through deployment and ongoing optimization. As enterprise AI adoption continues to mature, integrated platforms capable of combining proprietary enterprise data with continuous human-guided model improvement are likely to become an increasingly important part of how organizations move AI from experimentation into production at scale.