The insurance industry is past the question of whether to apply AI to claims; the conversation has shifted to how fast and how deep. Generative AI alone was estimated by Bain ¹ to unlock a global $100 billion benefit opportunity for P&C insurers, through a 20–25% reduction in loss-adjustment expense and a 30–50% reduction in claims leakage. Yet FNOL automation remains shallow for most carriers, who typically run four to seven disconnected systems across policy, claims, billing, underwriting and fraud — integration drag that limits how far AI can actually reach.

I have written earlier articles on the preference for insurers to be able to deploy non-native AI versions or AI-Native versions in production environments in order to match their appetite for experimenting. When compliance is at stake, take the classic ‘The Tortoise and the Hare’ fable by Aesop to heart.

Over the last two years⁴ I have always ranked 5 Sigma and Snapsheet as two of the best claims platform technologies and have used them as a benchmark in this article.

5 Sigma and Snapsheet have each pushed the market past spreadsheet-era claims handling, and each has genuine strengths — 5 Sigma's Clive as an AI adjuster persona, Snapsheet's no-code rules engine and payments automation. Neither, however, has solved straight-through processing for carriers whose distribution runs predominantly through brokers and MGAs. An independent comparison review² specifically finds that 5 Sigma lacks the regulatory reporting, subrogation workflow and jurisdictional rule engines that enterprise carriers need; Snapsheet's own product material claims broader multi-party and regulatory capability, but neither vendor publishes³ a governed, versioned model of delegated authority as a distinct platform feature — the whitespace this brief targets. Zoomed out, the wider vendor landscape — full-core incumbents, specialist claims platforms, data/estimation providers, and horizontal enterprise tools — is bifurcating around a single question: which platforms offer genuine multi-model AI orchestration inside a governed, auditable architecture, rather than selling isolated point solutions.

This brief sets out the case for a composable claims platform: a headless, API-first core with no-code configurability for carriers, MGAs and brokers, a three-layer AI architecture (predictive, generative, agentic), and delegated authority modelled as a governed, first-class object — built specifically to close the gaps 5 Sigma and Snapsheet currently leave open, while remaining interoperable with the Tier 1 and Tier 3 ecosystem rather than trying to replace it.

A majority of claims professionals now cite processing efficiency and cycle-time reduction as their core transformation goals, and after a year of proof-of-concept work in 2025, insurers are moving to scale AI fully into production in 2026. The scale of the prize is what is drawing investment across the ecosystem: Bain & Company estimates that generative AI could reduce P&C loss-adjusting expenses by 20–25% and claims leakage by 30–50%, creating more than $100 billion in benefits for insurers and customers globally — not a US-only figure, and not a result anyone has yet banked, but a projection of what full-scale adoption could unlock.

The gap between that projection and current reality is itself the strategic opportunity. Bain's own follow-up survey of 160 global insurers found that while 78% of P&C insurers have adopted generative AI in some form, only 4% have scaled it meaningfully across their claims operations — most deployments remain piecemeal, applied to isolated tasks like fraud detection, document summarisation or customer communication rather than threaded through the full claims process. Insurers that took a genuine end-to-end approach saw materially different results: a 35% productivity boost and homeowners' claims processing times cut in half, versus the marginal gains typical of task-level automation. That distinction — bolt-on AI versus full-lifecycle AI — is the fault line this brief argues a new platform should be built around.

Underneath that adoption gap sits a structural one: most carriers currently run four to seven disconnected systems spanning policy administration, claims, billing, underwriting, fraud detection and third-party vendor applications, creating integration drag that caps how far any single AI initiative can reach, however capable the model behind it. The real opportunity lies in AI threading through the entire claims lifecycle, not sitting at any one point in it.

Digital FNOL is only one of at least eight points in the claims lifecycle where AI is already delivering measurable results. A platform strategy built around a single intervention point — however well executed — will always underperform one designed to thread AI through the full lifecycle.

Both 5 Sigma and Snapsheet have industrialised digital capture and workflow-routing of a claim. Independent market analysis puts typical straight-through processing rates across the sector at only 15–30% of claim volume overall — rising to 70–90% for simple claims on the most mature AI-powered systems, but still leaving a large share requiring adjuster review. Two structural issues explain the gap:

• Distribution mismatch. Where most premium is broker- or MGA-placed, the carrier's own digital front door is often the wrong door — the broker's own system, or the client's, is the actual point of first contact. Neither Five Sigma nor Snapsheet publishes a broker-embedding layer as a core part of its architecture; both are sold and deployed primarily as carrier-, MGA- or TPA-facing systems.
• Authority, not automation, is the bottleneck downstream of FNOL. Reserving, coverage confirmation, settlement and payment all depend on who is allowed to bind what, on whose behalf, up to what limit. An independent comparison review specifically finds that Five Sigma lacks the regulatory reporting, subrogation workflow and jurisdictional rule engines enterprise carriers need; Snapsheet markets broader multi-party and multi-line capability, but — like Five Sigma — does not publish a governed, versioned model of delegated authority as a distinct platform feature.

A platform that closes this gap needs to treat delegated authority — broker binding limits, MGA settlement authority, TPA mandates — as a governed, versioned object inside the core claims data model, not a permissions table bolted on as an afterthought.

The tension between no-code configurability and robust multi-party API connectivity is a false trade-off if the architecture separates the two correctly — and it is also where the incumbents currently split the difference rather than resolving it.

Headless, API-first core.  Claims logic, data model and orchestration are exposed via documented APIs and webhooks, so the platform plugs into broker systems, MGA binder platforms, telematics feeds, repair networks, medical panels, payment rails and legacy policy administration systems without bespoke integration work for every new partner. Incumbents claims software platforms offer API layers, but do not publish broker-side embedding as a first-class integration pattern — a specific, provable gap to design against.

No-code layer over a governed core.  Carriers, MGAs and brokers design and iterate workflows, decision rules and document templates visually. Snapsheet's rules/decision engine is a genuine strength and the clearest functional benchmark to match or exceed; Five Sigma's configurability is adequate for standard commercial lines but not for the deeper regulatory and jurisdictional logic enterprise carriers require. The opportunity is a no-code layer that goes further than Snapsheet's assignment/SLA rules into governed authority and reserve logic, without requiring custom development.

A critical insight echoed across the vendor landscape — Spear Technologies makes this point explicitly — is that no single AI model can meet every need. Leading platforms are adopting multi-model architectures across three layers, and a composable platform should be built around the same separation of concerns:

• Predictive AI — machine-learning models for fraud scoring, reserve prediction, STP routing and litigation propensity.
• Generative AI — document drafting, summarisation, coverage interpretation and customer communications.
• Agentic AI — orchestrating multi-step workflows autonomously, escalating to humans at defined thresholds, operating across systems in real time — the model Allianz's “Project Nemo” (launched in Australia in 2025) demonstrates in production, using specialised, task-oriented agents that plan, decide and collaborate on low-complexity, repetitive claims.

Every no-code change and every AI agent action should compile down into the same versioned, auditable objects the API layer uses, so agility and AI autonomy do not come at the cost of the regulatory depth that independent reviewers say both incumbents currently lack.

Positioning a new platform requires understanding the full ecosystem it will be evaluated against, not just its nearest competitors.

Both vendors are credible, well-funded, cloud-native platforms with real customer traction — 5 Sigma is positioned as the bridge between spreadsheet-based claims handling and a full Guidewire implementation, while Snapsheet counts 170+ customers including 16 of the top 20 P&C carriers. The comparison below identifies where a new entrant can differentiate rather than compete head-on.

What this means competitively

• 5 Sigma's edge is its AI adjuster persona (Clive) and speed of deployment; an independent review identifies its published gap as depth — regulatory reporting, subrogation, and jurisdictional rules — which matters most to carriers operating across multiple regulatory regimes.
• Snapsheet's edge is its no-code rules engine, payments automation and policyholder-facing mobile intake. Its own marketing claims broad multi-party and regulatory capability, so no equivalent independent “depth gap” finding exists for Snapsheet — but neither vendor's public material describes delegated authority (broker binding limits, MGA settlement authority, TPA mandates) as a governed, versioned object in its own right.
• This absence — confirmed by an adverse independent review for 5 Sigma, simply unpublished for Snapsheet — is the clearest whitespace: a platform that is simultaneously as configurable as Snapsheet, as AI-forward as 5 Sigma, and additionally provable on authority and jurisdictional compliance in a way neither currently demonstrates.

The strategic choice for insurers and TPAs is increasingly nuanced, and a new entrant needs a clear answer to where it sits among the paths carriers are already weighing:

• Embeddable FNOL and status APIs/widgets that brokers can drop into their own systems, closing the distribution-mismatch gap neither incumbent currently addresses as a core, published capability.
• Delegated Authority as a governed object: broker and MGA binding/claims authority modelled explicitly, versioned, and enforced at runtime — so a TPA can settle automatically within an agreed limit, with everything above it routed for carrier sign-off, and the carrier can demonstrate exactly why each automated settlement occurred.
• A no-code rule and workflow builder that matches Snapsheet's configurability for assignment, SLA and payout logic, and extends it into reserve authority and jurisdictional rules — the specific depth gap flagged in independent reviews of Five Sigma.
• AI copilots and agents (triage, damage assessment, fraud scoring, settlement recommendation) that operate strictly inside a defined mandate, matching Five Sigma's Clive and the sector's agentic direction (Shift Technology, Agentech, Project Nemo) on capability, while keeping every automated action bounded and auditable.

Claims AI cannot be designed in isolation. The coming years will see insurers coalesce AI transformation across underwriting, claims, policy servicing and customer experience, extending into actuarial analysis, compliance, finance, producer management and talent development. A composable claims platform needs to connect to:

• Underwriting systems — claims experience feeding back into pricing models
• Customer/CRM platforms — a single view of the policyholder across service and claims
• Finance and reserving — real-time reserve adequacy
• Reinsurance bordereau — automated cession and loss reporting
• Fraud intelligence networks — cross-carrier data consortia
• Regulatory reporting — IFRS 17, Solvency II, and state- or jurisdiction-level AI compliance regimes

The regulatory environment is sharpening in step with the technology. In the US, the NAIC has released an AI Principles framework addressing transparency, explainability and non-discrimination, with state-level rules on claims-specific AI still evolving; in the UK, the equivalent pressure comes through the FCA's Consumer Duty and SM&CR, which require carriers to evidence fair customer outcomes and maintain individual accountability for automated decisions. Either way, the direction of travel is the same: platforms will increasingly be judged on whether their AI is governed and auditable, not just on whether it is fast.

The market is bifurcating: carriers treating AI as a bolt-on to legacy processes are seeing modest FNOL gains, while those embedding AI deeply across the claims lifecycle — connecting assessment, adjudication, fraud, subrogation and CAT response into a coherent architecture — are achieving structural cost and experience advantages. The vendors that will dominate are those offering genuine multi-model orchestration (predictive, generative and agentic) within governed, auditable workflows, rather than selling individual point solutions.

• Genuine STP for broker-led books — embeddable FNOL and governed delegated authority remove the two structural blockers that neither Five Sigma nor Snapsheet currently solves as a core, published feature.
• Match-and-exceed configurability — no-code process design at least as capable as Snapsheet's rules engine, extended into reserve and jurisdictional logic that independent reviews say Five Sigma lacks.
• Full-lifecycle AI, not point-solution AI — predictive, generative and agentic capability threaded from pre-loss monitoring through subrogation and CAT response, matching the direction the whole vendor landscape (Duck Creek, Shift Technology, Allianz's Project Nemo) is already moving.
• Interoperable by design — built to sit alongside Tier 1 core platforms and Tier 3 data/estimation providers rather than compete for their role, so carriers are not asked to make an all-or-nothing platform bet.
• For insurers and TPAs evaluating options, the critical questions remain: how open is the API architecture for ecosystem integration, how mature is the vendor's agentic AI roadmap, and how will the platform scale during the increasingly frequent CAT surge periods that now define the industry's most critical operational moments.

Sources:

1. Bain & Company, “The $100 Billion Opportunity for Generative AI in P&C Claims Handling” and related Bain survey of 160 global insurers;
2. Independent comparison review (Lido.app, “Best Insurance Claims Processing Software,” 2026) for the 5 Sigma depth-gap finding; other independent market analysis (InsurAItools, HFS Research, Everest Group); and industry landscape research current as of mid-2026. Prepared as a strategic positioning brief for discussion purposes.
3. Vendor-published product material (fivesigmalabs.com, Snapsheetclaims.com
4. Insurtechworld.org 30^th June 2026 and 26^th May 2026 for latest articles
5. Insurtechworld.org 14^th April 2026