Turning data into decisions insurers can trust

When a claim is filed, critical decisions happen in an instant. A single email, a scanned PDF, or a few keystrokes can determine whether a payout is fast, a risk is flagged, or a customer’s trust is preserved. Too often, that information is stuck in silos or slowed by outdated systems. What if those scattered fragments could speak to each other, reason in real time, and anticipate needs?

When a claim is filed, critical decisions happen in an instant. A single email, a scanned PDF, or a few keystrokes can determine whether a payout is fast, a risk is flagged, or a customer’s trust is preserved. Too often, that information is stuck in silos or slowed by outdated systems. What if those scattered fragments could speak to each other, reason in real time, and anticipate needs?

For decades, data in insurance was treated as a ledger. It was static and reactive. Designed for storage and retrieval. But over the past year, a new paradigm has emerged, one where data drives decisions in real time.

Henry Peter, CTO at Ushur, describes the change as moving to a powerful and flexible agentic solution. The firm is moving away from reliance on pre-built workflows, and towards intelligently pre-empting what needs to happen next.

He explains, “Over the past few years, Ushur’s use of data has evolved from structured, reactive workflows to adaptive, outcome-driven intelligence—enabled by advancements in domain-specific AI and multi-agent orchestration.”

Three forces are driving the shift. First, technology: advanced intelligence systems can process both structured and unstructured information, from documents to conversations, while applying context-aware reasoning.

Second, customer expectations. People now demand services that anticipate their needs, resolve issues proactively, and simplify interactions.

Finally, evolving rules and compliance standards are redefining how information is collected, stored, and used, making governance an integral part of every process.

The power of smarter data

The real competitive edge lies not in having more data but in having smarter data. Peter points to two domains making the biggest difference.

The first is vertical, domain-enriched interaction data. Ushur’s models are trained on years of insurance-specific workflows and compliance rules, giving its AI agents a kind of native fluency in industry context.

The second is unstructured communications. Free-text emails, handwritten forms, and scanned documents are converted into actionable intelligence through optical character recognition and enrichment pipelines.

Peter explains the power of this combination, saying, “These sources power AI agents that can process and orchestrate across multiple domains—intake, underwriting, billing, claims—at scale, with precision and compliance baked in.”

Turning fragmented data into actionable intelligence

Unlocking that value is far from automatic. Organisational silos, legacy technology, and regulatory friction remain major obstacles. Customer journeys often span underwriting, billing, and claims, yet no single team owns the whole arc. Legacy systems resist integration with modern tools, and complicated compliance rules can slow every step of data capture and exchange.

Peter takes a pragmatic perspective on these challenges, “While the barriers are real, Ushur turns them into building blocks. By allowing insurers to incrementally modernise and scale at their own pace, Ushur helps extract value from data without compromising security, compliance, or operational stability.”

Ushur’s multi-agent orchestration coordinates across functions without requiring companies to reorganise.

Its integration-first design overlays intelligence on existing systems, modernising without disruption. Its platform embeds auditability and privacy into every layer, making compliance a foundation, not a roadblock.

Other leaders in the space echo the sentiment, highlighting additional layers of complexity. Paulo Ferreira, CTO of KYND, adds a complementary perspective, focusing on data usability across the enterprise.

“One of the biggest barriers isn’t just the data itself, it’s making sense of it in a way that’s actually usable for decision-making,” he says. “A lot of the time, insurers are faced with fragmented, inconsistent, or overly technical cyber data that doesn’t translate easily into underwriting action.

KYND, instead, focuses on turning that complexity into clear, usable intelligence that drives confident decisions.

Ferreira emphasises that trust and explainability are central to turning data into actionable insights. “If underwriters can’t see where data comes from or how it’s been derived, they’re right to question it. Transparency and explainability are often missing, and without them, data becomes a blocker rather than a benefit.

Vikas Acharya, CEO of ChainThat, highlights the operational side of integrating data across workflows. He describes a shift over the past year toward treating data as a strategic asset, integrated into every stage of the policy lifecycle.

“In the past year, we’ve moved decisively towards data as a strategic asset, designed into every stage of the policy lifecycle, from product configuration to claims, rather than captured as an afterthought,” he explains.

“This evolution means data is no longer passive. It’s actively used in underwriting guidelines, workflow automation, and decision-making within BPA. For example, sanctions data can instantly move a case into referral, while hazard and valuation data pre-fill risk profiles for underwriters.”

Acharya further underscores how multiple, integrated sources give insurers a competitive edge, “Our competitive advantage comes from how we unify, govern, and operationalise multiple datasets within BPAensuring they are consistent, secure, and real-time across all workflows.”

The modern insurance promise

Fairness and trust remain central, especially as algorithms shape customer outcomes. Transparency is built into Ushur’s design. Every AI interaction can be traced and audited.

Guardrails enforce access controls and privacy standards, and pricing is modular and usage-based, making cost structures as explainable as the models themselves.

Peter explains, “Customers access the full Ushur platform and add products like Invisible App for secure app-like self-service, Ushur Hub for activity orchestration, or Ushur Intelligence for deploying agentic AI across use cases. Pricing scales with usage and specific AI capabilities, ensuring enterprises pay only for what they use while aligning cost with business impact.”

Ferreira reinforces the connection between data quality and actionable insights, “High-quality, broad and timely data has always been at the core of our business. The real value lies not just in having the data, but in the ability to identify and present the insights that matter, ensuring the right information to the right people, at the right time.”

Acharya expands on explainability, detailing the platform’s safeguards, “Every change to rating rules, workflows, or mappings is logged and attributable. BPA links each decision—pricing, referral, or claims—to the exact rule or data point that triggered it. The platform records when data was pulled, from which source, what specific data was retrieved, when it was retrieved, and which user initiated the pull—ensuring a full audit trail for compliance and operational transparency.”

Insurance has always been a business of promises: to cover, to protect, to restore. Today, those promises rest on two modern pillars: data and trust.

One without the other is fragile. Data without trust is noise, and trust without data is hollow.

Together they form the foundation on which insurers can deliver speed, fairness, and confidence at scale. The invisible fragments, the emails, the forms, the scattered keystrokes, become the bedrock of decisions insurers and their customers can believe in.

Read the daily FinTech news

Copyright © 2025 FinTech Global

Enjoying the stories?

Subscribe to our weekly InsurTech newsletter and get the latest industry news & research

Investors

The following investor(s) were tagged in this article.