Company
Marjan is an E&S retail broker that focused on forced-place coverage.
HQ
SF
Industry
Insurance
INTRO
INTRO
Marjan is a family-run retail insurance broker started in the 90s by Mike Van Eerden and today run by Mike and his daughter, Jane. The company operates in the niche world of forced-place coverage, working with large commercial real estate lenders to ensure properties remain properly covered when borrowers fail to maintain sufficient insurance.
Mike and Jane have deliberately kept their team small but two of their core processes, generating cost estimates and preparing monthly carrier reports, still required days of manual work across spreadsheets, emails, policy documents, and historical reports, which meant they would need to hire as volume increased.
Marjan had already tried automating both of these processes with an AI development shop but the system was unreliable, slow, and struggled with the dozens of rules and edge cases involved in lender-placed insurance.
PART I
Turning lender requests into cost estimates
When one of Marjan’s lender clients needs to understand the cost of lender-placed coverage, they send Marjan a spreadsheet containing the loans and coverage they want priced. Marjan then has to apply the appropriate rates, limits, dates, and coverage rules and return a detailed Excel cost estimate.
While the basic process is repeatable, the details make it difficult to automate. Coverage can be backdated, rates vary by geography and coverage type, borrowers may already have partial insurance, and liability and umbrella coverage can interact in ways that change the final premium.
Historically, Marjan relied on a VBA-driven process combined with manual adjustments and QA. Returning a cost estimate could take one to two days.
Velos automated the workflow from the incoming lender request through the finished Excel workbook, including Marjan’s existing calculations, business rules, and formatting.
Today, Marjan can return a cost estimate to its customers in under 10 minutes.
PART II
Automating monthly carrier reporting
Monthly reporting was an even larger bottleneck.
Each month, Marjan has to reconcile what coverage was actually placed and determine what should be billed or refunded. Reports can include new placements, cancellations, backdated coverage, refunds, and changes with different effective dates.
Doing this correctly also requires understanding what happened in previous months. A refund might relate to coverage placed months earlier, while a new request may overlap with coverage already on the books.
Previously, Marjan’s team manually carried forward prior reports, reviewed new lender information and email threads, determined what changed, performed the calculations, and created the Excel tabs sent to its insurance partner.
The process took roughly three full days every month.
Velos automated the workflow while retaining the historical context needed to handle backdating, refunds, duplicate coverage, and other exceptions.
Today, Jane and her team are able to complete their monthly carrier reporting in under two hours as opposed to three days.
Outcomes:
Time savings: 30 hours saved each month on monthly reporting
Faster cost estimate: Able to return a cost-estimate to a client within an hour
Cost savings: No longer needed to hire more a full-time analyst
