Underwriting leaders face unprecedented technology choice: workbenches, intake accelerators, AI extraction, rules engines, workflow platforms, or assistant layers. The promise is faster turnaround, better risk selection, and less rekeying. The reality, however, is more nuanced.

After watching multiple underwriting transformations succeed, and others stall, I’ve reached a simple conclusion: there is no silver bullet vendor. Most credible solutions now rely on similar cloud services and patterns. Differences exist (UX, configurability, integration, analytics maturity), but in commercial and specialty lines, implementation quality is almost always the decisive factor. Success depends on fit to your operating model, data integrity, and how precisely appetite and process are translated into executable design.

Different systems – similar capabilities

Many of today’s underwriting platforms are built from similar technological components. They typically rely on document-reading tools, search and indexing engines, advanced analytics environments, and AI capabilities powered by large language models. This does not mean that all products are the same. However, it does mean that insurers should look beyond marketing claims and focus on what truly differentiates one solution from another.

The real question is whether the system removes your actual operational bottlenecks. Does it genuinely improve how submissions are received, triaged, enriched with data, evaluated, quoted, referred, documented, and audited? Or does it simply digitise existing inefficiencies?

Flexibility is equally important. Can the solution reflect your underwriting appetite and authority structure without requiring years of complex configuration? A system that requires excessive customisation often ends up recreating the very rigidity that transformation was intended to eliminate.

Control and transparency also matter. Strong audit trails and decision traceability are no longer optional in an increasingly regulated and data-driven environment.

Finally, integration is critical. An underwriting workbench does not operate in isolation. It must connect cleanly with policy administration systems, pricing engines, claims platforms, and document management systems. Without that integration, even the most sophisticated front-end solution will struggle to deliver real value.

No two specialty carriers operate the same way. Tools may be similar; implementations never are. That is why SMEs, who understand insurance logic and integration/data implications, are the scarcest resource. They determine whether transformation becomes genuine improvement or “replatforming with the same pain.”

Customisation is unavoidable

In commercial/specialty underwriting, some customisation is always required. The real question is where and why. My advice would be:

• Differentiate on appetite, authority, referral logic, risk signals, audit.
• Standardise workflow patterns, document capture, integration plumbing.

Excessive customisation may feel like control, but it quietly creates long-term drag. Every deviation from the standard increases complexity, makes upgrades harder, testing more expensive, and change slower. Customisation should be treated like a serious investment decision – justified, deliberate, and sustainable, not a default reaction.

Building a fully in-house underwriting platform is rarely the right answer unless your process is genuinely unique. Maintaining such a system becomes a permanent obligation: ongoing enhancements, security, regulatory updates, and continuous improvements. Even well-executed bespoke builds often come with a high hidden cost in time, focus, and capital that could have been invested elsewhere.

Standard solutions will never be a perfect fit. But if they allow you to go live in months rather than years and remove real operational constraints, speed and focus often outweigh the compromises. A practical principle emerges use market-standard tools for common functions and concentrate your energy on what truly sets your underwriting apart.

When a targeted inhouse component makes sense

A narrow, well scoped inhouse component can be a smart bridge, especially when budgets are tight or a major PAS program is upcoming. If you know the bottlenecks that dominate cycle time (e.g. ingestion, triage/routing), a lightweight, targeted solution delivered as a service, incorporated into your current workflow can relieve pressure quickly without committing to a full platform rebuild. If you know exactly what problem you are solving and keep scope tight, targeted builds can deliver fast, measurable value.

The critical caveat: small builds can become shadow platforms

The most common failure mode is scope creep: triage becomes “a bit of workflow,” then “a few screens,” then “some pricing,” then “we’ve basically built a mini-workbench.” At that point you inherit the same product management burden you were trying to avoid, just with fewer controls and higher keyperson risk.

When considering an inhouse component, be disciplined about:

• Ownership and longevity: who runs it when engineers change
• Operational resilience: uptime, monitoring, error rates
• Security, compliance and audit: especially for AI supported actions
• Data quality and feedback loops: without trust, adoption collapses
• Model and rule drift: appetites evolve; accuracy decays without governance
• Total cost of ownership: the next 18 months matter more than the 30-day build

Closing thought

From my perspective “find the best tool” is the wrong goal. The right goal should be to remove the constraints that prevent fast, well governed underwriting decisions. Sometimes that means buying a workbench and implementing it well. Sometimes it means a targeted component to relieve acute pressure while larger programs mature. What consistently works is clarity of outcomes, controlled scope, strong SMEs, and respect for underwriting realities: variability, exceptions, explainability.

Technology amplifies judgment – but only when the operating model makes that judgment repeatable.

Author of the article: Jakub Śliwiński, Head of Underwriting at Sollers Consulting