Understanding the challenge
The brokerage ran on re-keying. Every new-business submission, renewal, and certificate of insurance meant the same account data was typed and retyped across carrier portals, the broker management system, spreadsheets, and email, by hand, by some of the firm's most experienced people. Submissions that should have gone out same-day sat for two or three days in a queue of manual steps. "My account managers were hired to advise clients and place risk, not to copy policy numbers between six screens," the operations director said. Renewals slipped because reminders depended on someone remembering, and approvals waited in inboxes. The firm knew automation could help but had never mapped where the time actually went, and every past attempt had stalled on the sheer number of systems that did not talk to each other.
Our approach
We started our Workflow Automation engagement where it should start, with a workflow audit, mapping every high-volume process to find where work slowed down, duplicated, or leaned too hard on manual effort. Then we automated the bottlenecks first, connecting the systems the brokerage already ran on so work moved without manual handoffs:
• Submission and renewal intake that reads the incoming documents, extracts the account data, and moves it into the CRM automatically, using AI where the information arrived unstructured, so account managers stopped re-keying.
• Automated renewal reminders and follow-up emails, so nothing slipped because someone forgot.
• Approval and hand-off routing between service, production, and accounting, so files moved instead of sitting in an inbox.
• Lead routing and reporting automation that put new inquiries in front of the right producer and gave leadership a live view without anyone building a spreadsheet.
Technical innovation
The heart of the build was a set of connected automations linking the brokerage's CRM, forms, email, spreadsheets, and reporting systems into one flow, so a submission entered once moved through intake, data entry, routing, and reporting without a person re-typing it. Where documents arrived as unstructured PDFs, an AI step read and classified them before the automation took over, which is exactly the line to draw: simple automation for the repeatable steps, AI only where judgment or unstructured data is involved. Every automated step is logged, and exceptions route to a person, so the team keeps oversight of anything unusual.
Outcome
New-business submissions that used to take two or three days now go out the same day, roughly three times faster. Account managers got back about fifteen hours each per month that had gone to manual data entry, and manual re-keying across systems fell by around 70%. Renewal reminders stopped depending on memory, so fewer renewals slipped, and leadership finally had live reporting without anyone assembling it by hand. "The work still gets done carefully, it just does not eat my team's day anymore," the operations director said.
