The launch of Bevaya marks a pivotal moment in the insurance technology landscape, signaling a shift from niche AI experiments to enterprise‑wide intelligent automation. Roots Automation, Inc. has retired its legacy Roots brand in favor of Bevaya, positioning the new platform as the flagship solution for carriers, brokers, and third‑party administrators seeking to embed agentic AI into core workflows. This rebranding is more than a cosmetic change; it reflects seven years of field‑tested learning, over a hundred production deployments, and a deepening understanding of where AI can truly augment human expertise. For industry leaders watching the slow creep of digital transformation, Bevaya offers a concrete pathway to move beyond pilot purgatory and achieve measurable operational impact at scale.
Founded in 2018 by two former AIG executives, Roots Automation began with a simple observation: talented insurance professionals were spending excessive time on repetitive, rules‑based tasks that technology could handle more efficiently. Over the subsequent seven years, the company refined its approach, moving from isolated proof‑of‑concepts to robust, scalable AI agents that now serve some of the largest players in the market. The journey from early adopters to enterprise‑scale partners has been documented through 115+ production implementations, spanning three of the top five property‑and‑casualty carriers, three of the top ten brokers, and three of the top twenty TPAs. This track record provides a solid foundation for Bevaya’s claim that it delivers AI that not only works in theory but thrives under the pressure of real‑world insurance operations.
Bevaya’s technical architecture introduces a unified environment where users can design, deploy, and govern AI agents across the entire insurance lifecycle—underwriting, claims, and policy servicing—without juggling disparate tools or bespoke integrations. The platform’s redesigned user interface emphasizes clarity and collaboration, allowing business experts, data scientists, and IT teams to co‑create agents on a shared Workflow Canvas. By centralizing governance, Bevaya addresses a common pain point in AI adoption: the difficulty of maintaining oversight, version control, and compliance as models proliferate across departments. This holistic approach aims to reduce the friction that often stalls enterprise AI initiatives and accelerate time‑to‑value for new use cases.
At the heart of Bevaya lies InsurGPT™, an ensemble of specialized language models trained on more than 300 million proprietary insurance documents. Unlike generic large‑language models that require extensive prompt engineering to grasp domain nuances, InsurGPT™ has internalized the specific jargon, procedural logic, and contextual subtleties that define insurance work. This deep domain fluency enables Bevaya’s AI agents to perform complex tasks such as triaging incoming claims, analyzing coverage applicability, rating risks, and recommending next‑step actions with a level of precision that mirrors seasoned underwriters or claims adjusters. The result is a technology that speaks the language of insurance fluently, reducing the need for constant human intervention while preserving the judgment that only experienced professionals can provide.
Chaz Perera, CEO and co‑founder of Bevaya, articulated the founding motivation during the launch event, recalling his years at AIG watching skilled employees bogged down by manual, repetitive work. He emphasized that the original mission—to replace drudgery with intelligent automation—has been validated by seven years of field experience, 115+ deployments, and over $100 million in realized customer value in 2025 alone. Perera’s statement underscores a broader industry truth: the most successful AI implementations are those that start with a clear understanding of the human pain points they aim to alleviate. Bevaya’s positioning as a platform built by insurance veterans for insurance professionals reinforces its commitment to delivering tools that respect the expertise of the workforce while augmenting their capabilities.
Bevaya’s performance metrics highlight its readiness for production‑scale deployment. Customers report capacity gains of three to four times, meaning that a single AI agent can handle the workload of multiple human associates on tasks like data extraction, initial assessment, and routine decision‑making. Accuracy rates exceed 98% on critical work, a figure that speaks to the rigor of the model training and validation processes employed by Roots Automation. Moreover, the platform promises measurable return on investment within the first year, a compelling proposition for CFOs and operational leaders who demand tangible financial outcomes from technology investments. These benchmarks are not theoretical; they are drawn from live environments where Bevaya agents are already processing millions of transactions annually.
Situating Bevaya within the broader market context reveals why its arrival is timely. While many insurers have experimented with AI chatbots, robotic process automation, or generic machine learning platforms, few have found a solution that balances deep domain specificity with enterprise‑grade scalability. Horizontal AI providers often struggle with the nuanced language and regulatory complexity of insurance, leading to brittle models that require constant retraining. Conversely, legacy rule‑based systems lack the adaptability needed to evolve with changing products and risk landscapes. Bevaya attempts to bridge this gap by combining specialized AI models with a configurable workflow engine, offering a middle path that leverages the strengths of both approaches while mitigating their weaknesses.
For carriers looking to adopt Bevaya, practical implementation begins with a clear definition of the target workflow—whether it be claims first notice of loss triage, underwriting risk selection, or policy endorsement processing. The Workflow Canvas enables subject matter experts to map out each step, decision point, and data requirement alongside AI specialists, ensuring that the resulting agent reflects real‑world operational logic. Bevaya’s AI Assistant can suggest model configurations, data mappings, and rule sets based on patterns observed in similar deployments, accelerating the design phase. Importantly, the platform supports iterative testing, allowing teams to run agents in shadow mode, compare outputs against human decisions, and refine performance before going live.
Brokers and third‑party administrators can leverage Bevaya to address pain points that are often overlooked in carrier‑centric discussions. For example, AI agents can automate the intake and preliminary analysis of broker submissions, flagging missing information, suggesting appropriate carriers based on risk appetite, and generating preliminary quotes. In the claims arena, Bevaya can assist with salvage and subrogation identification, reserve setting, and customer communication drafting, freeing adjusters to focus on negotiation and complex judgment calls. By standardizing these routine yet critical steps, brokerages and TPAs can achieve greater consistency, reduce errors, and improve service levels to their own clients—ultimately strengthening their competitive positioning in a crowded marketplace.
Governance and risk management are integral to Bevaya’s design, recognizing that insurance AI must operate within strict regulatory frameworks and ethical guidelines. The platform includes built‑in capabilities for model explainability, audit trails, and bias detection, helping organizations satisfy requirements from regulators such as the NAIC and state insurance departments. Administrators can set policies governing data access, model versioning, and human‑in‑the‑loop thresholds, ensuring that AI agents remain accountable and transparent. Moreover, Bevaya supports continuous learning loops where agent performance is monitored, and retraining triggers are based on predefined performance drift metrics, thus maintaining accuracy over time without introducing uncontrolled variability.
From a financial perspective, the adoption of Bevaya translates into concrete cost savings and revenue opportunities. By offloading repetitive tasks to AI agents, organizations can redeploy skilled staff toward higher‑value activities such as complex risk analysis, customer relationship management, and strategic product development. The capacity gains of three to four times effectively allow insurers to handle growth volumes without proportional increases in headcount, a significant advantage in a market where talent scarcity and rising labor costs are persistent challenges. Early adopters have reported reduced loss adjustment expenses, faster cycle times, and improved net promoter scores, all of which contribute to a stronger combined ratio and enhanced shareholder value.
To harness the full potential of Bevaya, insurers should follow a structured, phased approach. First, conduct an internal audit to identify high‑volume, rule‑intensive processes that are prime candidates for automation—think claims triage, policy endorsements, or broker submission intake. Second, assemble a cross‑functional team that includes business process owners, IT architects, data scientists, and change‑management specialists to co‑design the AI agent using Bevaya’s Workflow Canvas. Third, launch a limited‑scope pilot in shadow mode, measuring key performance indicators such as accuracy, processing time, and user satisfaction against a human baseline. Fourth, based on pilot results, refine the agent’s logic, expand its scope, and plan a phased rollout across additional lines of business or geographic regions. Finally, invest in ongoing training and communication to ensure that staff view AI as a collaborator rather than a threat, fostering a culture of continuous improvement and innovation.