Certificate and renewal automation, built inside Applied Epic or AMS360. Certificate requests arrive by email in every format, and someone checks each one against the policy. We put a senior engineer with your account managers, in your AMS, to decide what to buy and to automate that work in the systems you already run.
Bring a week of certificate requests from the service inbox.
Published results in insurance brokerages
Two large brokers published what certificate automation did, and how long accuracy took.
A leading U.S. insurance brokerage cut certificate-of-insurance processing time by 50% across about 500,000 email requests a year.
Accuracy was about 50% at go-live and reached 80% within three months, after corrected output was fed back into the models. EXL also ran training and change management.
Average handling time per certificate request fell from 41 to about 21.5 minutes at a global broker receiving about 2,500 requests a day.
Models configured for the broker's documents, API integrations into its systems, and a single validation screen in place of 132 manual ones.
Figures as reported by the vendor or customer at the linked source. They are not OpenNash results.
Where our engineer comes in
Accuracy at the U.S. broker started near 50% and took three months of corrections to reach 80%. That stretch between go-live and useful is where most brokerage AI projects stall. Our engineer works through it with you, with an AMS login, a desk next to the service lead, and test cases your account managers write.
| Decision | What we look at | Outcome |
|---|---|---|
| Which workflow goes first | Certificate requests, renewals, and policy comparisons per week, with time per item. | Usually certificates. The volume is high and the rules are clear. |
| Buy or build | Your AMS vendor's AI add-ons and newer brokerage tools, tested on your real requests and priced at your volume. | Buy where it covers the workflow. Build where your rules or carriers make it a poor fit. |
| What the model decides about coverage | Additional insured, waiver, and primary-noncontributory wording. | Nothing. The workflow checks policy fields against agency rules, and a licensed person handles every exception. |
| How carrier portals are handled | Which carriers have an API or download feed and which only have a website. | Logged browser steps where there is no API. Errors go to a person. |
Why forward-deployed engineers
Built for their largest accounts, deploying their own models. OpenAI's deployment company launched with more than $4B. ↗
Configure their own product, priced per seat or per resolution.
A small senior team, ex-Palantir, Meta, and Databricks. Any model, any vendor. Free pilot, then a flat monthly fee.
Even OpenAI closes the gap with engineers in the room. Source: OpenAI, What FDE means at OpenAI, Oct 2026.
The first eight weeks
Written down with your team in week 1. The pilot is free, and it has to pass these before anything goes further.
Example acceptance tests, not results from a customer deployment. Yours are written with your team.
Systems we work in
Rules from day one
FAQ
No. The workflow reads from and writes to the AMS you already run, and the AMS stays the record.
Your account managers. In the pilot every certificate is a draft. After launch, routine certificates can go out under rules you approve, and anything unusual still goes to a person.
Coverage checks run as tested rules on policy fields, not model judgment, and anything outside them goes to a licensed person.
Why you need one
A senior engineer joins your team on a laptop you issue, tests the options on your data, and puts one insurance brokerages workflow into production in 4 to 8 weeks. It starts with a free 14-day pilot judged against test cases you agree in writing.
After launch, we run and improve the workflow on a flat monthly fee, or your team takes it over. The code, tests, and runbook are yours either way.
Bring a week of certificate requests from the service inbox.