Resources
Blog
January 9, 2024
xx min read

Training Loss Runs

Convr takes P&C insurance customers’ unstructured claims data and turns it into something that underwriting teams can use to make better decisions, faster.

Assessing the risk of an insured accurately is of utmost importance for underwriting teams to keep their company’s books balanced. Often times insurance organizations will pull loss run reports on their policyholders to assess how loss-prone they’ve been over a specific time period.
Insurance providers can get that complete look by reviewing the list of claims generated by an insured, in what’s known as a loss run. Loss run reports give an insurance company a consistent look at their insureds. The report includes information such as the date and description of a loss or incident, as well as the amount that was paid out to the insured by the company following such loss. And the report may yield results showcasing multiple losses.
Digging into the balance sheets of an insured on a routine basis allows insurance providers the opportunity to review the accounting and ensure their premium is adequate. But the process can be tedious. Oftentimes insurance organizations will look to offshore teams to pull loss run reports on their customers to get this information and it can take days to weeks to get the information back for review.

\n

Loss runs are essential for insurance underwriters and brokers to fully assess the risk associated with insuring a particular business. By reviewing the claims history, insurers can evaluate the likelihood of future claims and determine appropriate pricing for the insurance policy. But days to weeks to get a complete picture on an insured in today’s times? Yikes! This isn’t the Stone Age.

\n

That’s why more companies are asking about the loss run extraction capabilities of Convr. Convr AI plays a significant role in helping insurance companies by improving efficiency, accuracy, and decision-making processes. Underwriting teams can utilize Convr’s Intake AI product to ensure they’re extracting all the data pertinent to a risk to review the risk as completely and accurately as possible, including loss runs. Convr takes our customers’ unstructured data and turns it into something that underwriting teams can use to make better decisions, faster.

\n

How it works for loss runs:

\n

Customers send Convr loss run documents via an API, email attachment, or manually drag and drop files from their desktop or other device connected to the Convr platform, then Convr takes the documents and identifies the data within it and converts it from unstructured or partially unstructured data and structures it into a structured file.
Customers can also edit the extraction—via human-in-the-loop (HITL). Human-in-the-loop can help the customer meet any specific requirements or help with any challenges that were experienced during the extraction process by organizing and cleaning up any final details. Convr has a team of experts on standby for this or some customers use their own teams.
The results?
The results are clear cut and straightforward; the data we provide is easy to understand and digest and improves speed to quote for customers. There’s an entire user interface that allows the user to view the entire source document. Users can actually see what the original data was and see a side-by-side table view to compare it to the original data and contrast what was extracted from the loss run.
The outcome takes minutes, not days or weeks, and it happens here in the U.S., during regular business hours, not offshore. So the entire process unfolds in the middle of your workday, enabling your underwriting teams to carry on with business as usual with minimal delays or disruptions. How's that for efficiency?
And our successes with loss runs are nothing new. Convr has been continually training our models since we started eight years ago under the name DataCubes. We’ve processed over 10 million loss runs since then and continue to support larger and mid-size carriers as well as smaller, regional carriers with this important task. And there’s a feedback loop. The more loss runs we see, the more our model learns and is trained and the better it performs. Over time it becomes more accurate and faster as it gets more and more exposed to different documents and training.
Convr brings quick-time-to-value, but it’s really something you’ve got to see to believe. Sign up for a FREE demo now to see how you could realize ROI in as little as six to eight weeks.

blogs

Keep Reading

More articles on AI, underwriting and the future of commercial P&C.

XX MIN READ

Glean Insights on Hard-to-Find Small Businesses with Convr’s Biz Intel Feature

A huge portion of commercial property and casualty (P&C) insurance applicants barely exist online. Many small and mid-size commercial insureds (the bread and butter of commercial insurance underwriting) are nearly invisible online.

Think about it . . . landscapers, contractors, florists and more. The  food truck owners, small town auto mechanics and mom and pop shops . . . many don’t have:

  • a website
  • a strong social media presence
  • consistent business filings
  • complete insurance applications

Underwriting team members call this a low digital footprint risk and it’s a problem for them. When the submission comes in, they need to know if the business is real, if the owners do what they claim to do, and if the exposure is what the agent says it is.

But if the business has no digital presence, the underwriter is lost without their normal verification tools including website and online reviews, access to pertinent safety records and satellite exposure checks as well as prior filings.

That’s where Convr’s AI Underwriting Workbench shines. With our Biz Intel web search feature for low digital footprint companies, that hard to find information easily turns up for the underwriter within our underwriting platform. ‍

The Convr Underwriting Workbench’s Biz Intel can uncover:‍

1) Business Classification

2) Appetite relevant exposures

3) Number of employees

4) Revenue

It turns an unknown into a knowable risk, giving the underwriter the opportunity to decide whether or not to write the risk rather than to spend time investigating it further. It’s a shortcut for underwriting team members of all levels as they spend less time searching for the details that move the decision.‍

All in one place:‍

In the Convr AI Underwriting Workbench, every new submission with the web option enabled, runs Biz Intel and returns the results inline. The hard-to-find details land next to the submission you're working on, not three tabs away from it.‍

Why it matters:‍

Low-digital-footprint submissions take time that underwriters often can't justify spending. Enrichment surfaces the missing data automatically, so accounts that would have been deprioritized – or declined for lack of information – become writable.‍

Convr’s Biz Intel users get:

1) First-quote advantage: Brokers place business with the first to quote. If your underwriting team is out searching Google, the Secretary of State, checking maps and emailing questions – you could be missing out on deals. With Convr AI data enrichment, the data comes to the underwriter instead of the other way around – and the first quote is more often yours.‍

2) Reduced referral dependency: When reliable information on low digital footprint companies is available in the file, more submissions can be decided where they land. Junior underwriters escalate only the accounts that genuinely need a second set of eyes. Senior underwriters spend their time on the complex risks and judgment calls that actually require their experience – not on questions a richer file would have answered on its own. Across the team, consistency improves and cycle times tighten.‍

3) Greater portfolio profitability: This is the real return on investment. Commercial carriers rarely lose money on catastrophic risks. Instead, they lose money on thousands of slightly mispriced/misunderstood small and mid-size policies – and low-visibility insureds are exactly where this is most common.

Convr's AI Underwriting Workbench isn't a productivity system. It's a loss ratio control system. If thin-file submissions are costing your team time or premium, we should talk – visit us at convr.com today.

XX MIN READ

Convr Accelerates MGA Growth

From Intake automation efficiency to data modeling for hidden insights, Convr is helping Managing General Agents (MGAs) turn fragmented submission documents into structured, enriched data – accelerating clearance, rating, and quote times to unleash profitable growth.

Everyone knows the best models win at taming documents!\nOur Intake module ingests and enriches data from both structured and unstructured documents including PDFs, Excel and emails across commercial property and casualty (P&C) insurance asset types, including ACORDs, Inspection Forms, SOVs, Loss Runs, Schedules, and more.

Powered by Convr AI and the Risk Context Engine – a purpose-built commercial insurance ontology, knowledge graph, and semantic layer that powers a multi-line schema – transforming fragmented submissions into structured, decision-ready intelligence. By grounding every document, application and data source into a consistent schema, Convr Intake ensures contextually complete, consistent, and reliable data from the start. The result is faster processing, fewer manual touchpoints, and improved risk clarity for accelerated MGA Growth.

Through Risk 360 – a commercial insurance data lake comprised of the digital footprints of millions of businesses – Convr standardizes addresses, performs geo-coding, enriches submissions with CAT modeling codes, and adds property intelligence data such as distance to coast and other hazard indicators. The enrichment delivers a holistic decision-ready view of risk prepared for underwriting, rating, and carrier reporting.

By eliminating re-keying and reducing back-and-forth data gathering, submissions are ready to quote in less than 10 minutes! This is how our MGA customers underwrite smarter and faster to unlock substantial written-premium growth without adding to headcount.

If you’re exploring ways to scale faster with AI, better data and meaningful operational efficiency, Convr welcomes the opportunity to share how leading MGAs are using Convr today.\nJust reach out to Convr today to see how we can help!

XX MIN READ

Convr® Evolves Data Catalog for Faster, More Transparent Underwriting

Convr®, took a leap toward delivering the next big advancement in artificial intelligence (AI) underwriting through the enhanced performance and usability improvements of its Data Catalog within its commercial insurance underwriting workbench in 2026. Convr’s Data Catalog makes it easier for underwriting team members to discover the thousands of data sources that are compiled in Convr’s Risk 360 data lake.\nThe leading artificial intelligence (AI) company serving commercial insurance organizations with its underwriting workbench implemented a new view for the company’s Data Catalog which aligns with its commitment to transform the commercial insurance underwriting industry and enables frictionless underwriting.\nWithin the model, users can easily glean insights from its extensive list of nearly 3,000 external data sources. Underwriters can also search across all Convr data sources, filter through the list by location, and view detailed information about each source with greater clarity. This new format allows users to quickly understand what data is available and how frequently it is refreshed.\nWhat it shows:The Data Catalog displays detailed information about each data source, including:

  • Source name
  • Data type
  • Source update frequency
  • Convr update frequency
  • State
  • Last updated date

The new table format makes it much faster and more intuitive for users to locate the data they need. The clickable links are up-to-date avenues that serve as a streamlined source to greater information and transparency into submission data. Our customers have much to gain from this new functionality.\nTogether, these advancements within the platform's user interface mark a pivotal moment, advancing the industry toward a more intelligent and trustworthy underwriting process built on accessible, high-quality data.

Realize End-to-End Underwriting Excellence with Convr AI

Experience how commercial P&C insurance organizations benefit from submission through quote with a frictionless process enriched by AI decisioning, empowering them to make better decisions, faster.