Handling complex submissions with AI
5 minutes
AI transforms a highly manual, painful aspect of the underwriter's workflow: keying complex submissions into pricing models and raters.
Every underwriter can attest, handling complex submissions from an email inbox is painful. Picture the scene: a 50MB email lands in the inbox, packed with scattered spreadsheets, PDFs, and scanned slips. Somewhere in that mess is the information you need to price the risk… but first, you’ll need to hunt it down, validate it, and re-key it in manually.
This is the reality of handling and ingesting submissions in commercial and specialty insurance. But it’s 2025. Technology ought to be changing this process. Thankfully, now is that time.
Advances in AI, data preprocessing, and intelligent automation are transforming how insurers handle submissions. This technology promises to transform one of the most painful parts of the underwriting workflow into the most streamlined.
The anatomy of a complex submission
Submissions often arrive via an email delivered with a side of attachment soup. These could be oversized spreadsheets with sprawling Statements of Values (SOVs), interlinking spreadsheets, PDFs or scanned handwritten notes, littered with missing fields and riddled with inconsistencies.
Some SOVs are delivered in multiple languages. Others spread critical details across misaligned tabs or broken layouts. Without a way to reliably ingest, validate, and align this data, the downstream chaos is all but guaranteed.
It’s a highly manual process, which is often undertaken by underwriters, or outsourced to costly third-party contractors. The data handling work is repetitive, uninspiring, yet vital. It often involves:
Copy-pasting rows into raters
Mapping fields by hand
Converting files into standardised formats
Wrestling with “optimistic” Excel mappers that still demand significant human effort
Once done, that data often needs to be input into multiple systems – a workbench, a PAS, and beyond.
It's no surprise then that 86% of underwriters still spend more than two hours a day on manual data entry, and half say these inefficiencies directly limit their ability to underwrite effectively.
This is more than a time sink. This process introduces inconsistency and erodes trust in pricing decisions, precisely when confident, data-backed judgment matters most.
Data chaos is a real business risk
As explored in Unlocking the Power of Clean Data, underwriting teams tend to underestimate the impact poor-quality submission data has on overall business performance.
These key issues include:
Incomplete or inconsistent records
Poor formatting or ambiguous labels
Redundant or contradictory fields
Inability to cross-reference or connect submissions to historical data
These issues create friction, introduce bias, and delay the ability to make fast, accurate decisions. In today’s market, that delay, and inaccuracy, comes at a cost.
From our experience with leading insurers, faster submission turnaround times have a direct impact on bind ratios. Underwriters that respond quickly and confidently are more likely to win the business - and build trust with brokers and partners in the process.
Faster, better decision-making in an increasingly competitive environment
The forces shaping the industry are increasingly volatile and powerful. These new exposures, climate volatility, geopolitical tensions, and complex emerging liabilities are reshaping underwriting criteria. For underwriters, this expands the surface area of a risk, making them more complex to understand and price.
Yet many underwriters still operate in workflows bogged down by data entry, burdensome data cleansing and administrative drag.
In fact, our State of Pricing Report highlighted:
95% of US insurers say their pricing technology needs improvement
47% say their platforms haven’t delivered what was promised
More than 1/3 report losing business due to slow or inconsistent pricing processes
As hyperexponential CEO Amrit Santhirasenan notes, “The industry is shifting from a defensive posture toward one that embraces technology offensively - for decision-making, not just operational efficiency.”
Structured, trustworthy submission data isn’t a nice-to-have. It’s a critical foundation for the kind of confident, accurate decision-making that’s needed to keep pace with the complexity and speed of modern risk.
How AI is untangling complexity
This is exactly what hx Renew’s Data Ingestion Library tool is designed to do.
Built for underwriters working with messy, real-world submissions, our data ingestion library leverages large language models (LLMs) and intelligent preprocessing pipelines to ingest, parse, and align submission data, then bring it directly into a rater.
The tool handles both one-to-one mappings (e.g., addresses or limits) and semantic inference for more qualitative fields which typically prove more challenging to interpret - such as gauging how well a cyber applicant enforces MFA across their organisation, for example.
In a nutshell, our Data Ingestion Library tool can:
Ingest messy submissions, in a variety of formats
Understand formats, schemas, and semantic meaning
Map unstructured inputs (like "Timber" vs. "Lumber") to structured model fields
File pre-screening and shrinking to reduce compute costs and accelerate processing
Offer a clear review layer, so underwriters stay in control, through end-to-end visibility
Why hx AI delivers more than standalone ingestion tools
With hx Renew, ingestion is deeply integrated into your pricing workflow. The platform already understands your pricing model’s structure, and can automatically map incoming submission data to your schema. That means faster onboarding, fewer configuration headaches, and an immediate acceleration in speed to quote.
Standalone ingestion tools often require constant tuning as your models evolve. That’s not the case with hx Renew. Because the data ingestion library is schema-aware, it updates as your pricing models change. No need to reengineer ingestion rules every time you iterate. This flexibility empowers underwriters and actuaries to continuously develop models without breaking downstream workflows.
Critically, you’re not just capturing data to park it in a workbench. With hx Renew, that structured submission data is instantly usable inside your pricing tool - building a rich data asset over time and unlocking opportunities for performance benchmarking, triage automation, and portfolio insight.
AI that unlocks better underwriting
While many AI tools are sold as productivity boosters, the real benefit here is deeper:
Greater consistency across teams
Faster response times without compromise
More time spent reviewing submissions instead of wrangling data
In a world of rising risk complexity and increasing competitive pressure, underwriting teams need systems that enable them to move quickly and act decisively - with confidence in the numbers behind every decision.
Tools like our Data Ingestion Library help make that possible. Not just by saving time, but by improving the integrity of every step that follows.
Augmenting the underwriter with hx AI
Overambitious transformation often fails. The smart path? Solve specific problems like submission ingestion – a current bottleneck impacting efficiency and growth. Insurers who fix this with intelligent workflows will be more agile.
Crucially, realizing AI's benefits now demands the right technology infrastructure. Purpose-built platforms, unlike often limited self-built solutions, are key to unlocking immediate value in areas like submission processing and building a foundation for future AI innovation. Investing in these platforms is the most direct route to tangible results and lasting competitive advantage.
Want to see how AI-powered submission ingestion can help your underwriting team move faster and price smarter? Book a demo of hx Renew, and our data ingestion library, today.