Majesco

The platforms administer policy, billing, claims and benefits, and they would keep doing that with the models removed. AI sits inside them as Copilot, an assistant in the products, and as agents that take on bounded tasks such as intake, reconciliation, triage and payment resolution. Majesco brands the platform AI native, while the shape of the product is a core system with AI built into its workflows.

Agents carry out bounded tasks through planned steps and checkpoints, and in supported workflows users review, modify, approve or override recommendations and proposed actions. The confidence threshold, and what an agent such as the claim reopening agent may finish unattended, are not stated.

Life and health intake agents complete straight through work where confidence is high and return exceptions with source rationale and decision context, and the Fall '26 release puts staff and agents on the same case in a shared workspace.

Recommendations are traceable: in supported workflows a user sees the context, inputs, recommended action and activity history, and life and health intake agents return each exception with its source rationale and decision context. Accuracy figures, validation methods and model documentation are not published for Copilot or the agents. Telemetry, a model catalog and an agent and tool registry are part of the AI framework the Fall '26 release extends.

Majesco reports 176 customer go lives in fiscal 2025, 80 percent of customers on current product versions and 95 percent on its cloud, figures that count systems in production rather than contracts. eMaxx went live on the property and casualty policy and premium audit platforms, and Crum & Forster, Heritage Property & Casualty Insurance and 32BJ Benefit Funds are named customers.

Outcome figures such as claims triage falling from 70 minutes to 2 come from modeled benchmarks rather than a named customer. Gartner named Majesco a Leader for SaaS property and casualty core platforms in North America in 2025.

Majesco commits in general terms to building its AI with security, privacy, governance and human oversight in mind. Whether one insurer's records inform models or agents serving another, and what Azure OpenAI Service keeps from Copilot prompts, is not stated. The trust center lists encryption at rest and in transit, backups and data access controls.

The privacy policy does not apply where Majesco processes personal data on behalf of its customers, which leaves the insurance and retirement records the platforms hold outside it. A data processing agreement sits in the trust center behind an access request, so its terms are not public. Those records now include benefits, absence and retirement data alongside property, casualty and life books, and Copilot runs on Azure OpenAI Service beside whatever a user has on screen.

Majesco's trust center, run on SafeBase, lists SOC 1 Type 2 and SOC 2 Type 2 reports, ISO/IEC 27001:2022 certification, a penetration test report, and the Standardized Information Gathering and Consensus Assessments Initiative questionnaires. The reports and the data processing agreement are released on an access request rather than open to download.

Compliance work here is product capability, from regulatory updates in each release to the rules behind absence, claims and benefits administration, and Majesco states no regulatory position of its own. Its AI timeline lists AI regulatory compliance among its 2026 aims without naming a regime.

The trust center lists four privacy laws, from the California Consumer Privacy Act to the Health Insurance Portability and Accountability Act, as compliance items, which are laws the company addresses rather than certifications it holds.

Majesco publishes AI governance commitments built on human review points, traceable recommendations, roles, approvals and change history. Bias testing or a fairness evaluation is not published for any of the AI features. The gap matters most where models score risk: the Fall '26 Intelligent Underwriting Workbench applies automated risk scoring to group and voluntary benefits, and agents work in claims, including claim reopening.

QKS Group placed Majesco at the top of its 2026 AI maturity evaluations for property and casualty core platforms and life policy administration, an assessment of capability rather than of outcomes across groups of people.

Recommendations carry their context, inputs and activity history in supported workflows, and intake agents return exceptions with source rationale, which gives an insurer a trail to review a contested decision. Terms on who bears the loss when Copilot or an agent is wrong are not published, and policyholders, members and claimants have no stated route to see or contest the reasoning. That gap is widest on the claim reopening agent and in retirement administration, where an error reaches someone's further hearing on a loss or their retirement income.

Copilot was developed with Microsoft and runs on Azure OpenAI Service, and Amazon Web Services hosts the infrastructure. Which models the agents and the underwriting risk scoring run on is not stated, and a subprocessor list is not public. The AI framework includes a model catalog and an agent and tool registry.

The AI framework adds an application programming interface, an agent console software development kit, a Model Context Protocol server and agent to agent orchestration, and a partner ecosystem lists integrations. CoreConnect and Enterprise Rating link product setup to pricing and quote, and general liability audit workflows in Loss Control 360 connect back to Majesco Policy. The production record is a delivery one: 176 go lives in fiscal 2025 and 80 percent of customers on current product versions.

Regions, tenancy and data residency terms are not published, which matters for retirement and pension records kept for decades. Majesco delivers the platforms from its cloud, where 95 percent of customers run, with Amazon Web Services listed as infrastructure in its trust center and Copilot on Microsoft's Azure OpenAI Service.

The platforms are quoted privately, with no price, unit or tier published. A buyer can size the counterparty instead: revenue is about $500 million after the Vitech acquisition, and 80 percent of customers run current product versions, which bears on the cost of staying current after go live.

Property and casualty, life, annuity and health, and retirement and pension each have their own product lines and release material, and the Fall '26 release kits are published market by market. More than 275 of the 375 plus customers are property and casualty insurers, and more than 100 are life, health, retirement and pension providers. Buyers named include insurers, reinsurers, managing general agents, brokers, third party administrators and Taft Hartley funds.