Case Study
An Operating Platform for a Bilingual Insurance Brokerage
An insurance brokerage serving English- and Spanish-speaking markets was bleeding hours on three different problems and did not realize they were the same problem: slow lead conversion, manual certificate-of-insurance tracking, and zero content output. We delivered three connected systems on a shared data layer.
Industry
Insurance Brokerage
Markets
EN + ES
Timeline
5 Weeks
Active Policies
5,000+
The Challenge
The owner was personally chasing every English- and Spanish-speaking inbound lead through SMS, email, and WhatsApp, on a window that often closed within two hours. When a referral partner sent over a list of clients, the firm had no automated way to know whose certificate of insurance was about to expire, whose was already expired, or which broker last touched that account. And the bilingual market in their region was enormous, but their site had no blog, no SEO presence, and no engine to produce one without hiring a full marketing team.
The owner summed it up: "I was the chatbot. I was the COI tracker. I was the marketing team. The business was not going to scale past me until that changed."
What We Built
We delivered three connected systems on a shared data layer.
The Comms Service is a production Express and PostgreSQL backend running BullMQ for queueing and a custom-trained chatbot built on Claude. The chatbot speaks fluent English and Spanish, handles 34 baked-in FAQs, and runs 34-message activation, nurture, appointment, opt-out, and retention sequences across SMS and iMessage.
The COI Tracker ingests partner-provided client lists, flags expiring certificates, surfaces them by broker, and produces re-engagement workflows for the partner relationship.
The Bilingual Blog Engine is a Next.js content pipeline with daily automated publication, an insurance compliance filter so nothing risky goes live unattended, and full SEO scaffolding in both languages.
System 1
Bilingual Comms Service
Express + Postgres + BullMQ. Claude-powered chatbot with 34 baked-in FAQs across SMS and iMessage.
System 2
Lifecycle Sequences
34 messages: activation, nurture, appointment, opt-out, retention. Tracked end to end.
System 3
Automated COI Tracker
Ingests partner client lists. Flags expiring certificates by broker. Re-engagement workflows built in.
System 4
Bilingual Blog Engine
Daily Next.js content pipeline with insurance compliance filter. SEO scaffolding in EN and ES.
Sample bilingual chatbot conversation
Live transcript
Inbound, Spanish
Alessandra, ~3s response
Booking handed off
Results
First-response time to new leads collapsed from hours to under 60 seconds, in either language, 24/7. Certificate expiration visibility went from 'best guess' to a single dashboard with broker accountability. The blog ships fresh, compliance-checked bilingual content daily without a marketing hire. Referral partner engagement is now measurable, with each partner seeing their own funnel. The owner is no longer the bottleneck, and the brokers are free to do what brokers actually do: write policies and run renewals.
Why This Matters
Insurance brokerage is a relationship business with a deeply unsexy back office. The firms that survive consolidation are the ones who automate the unsexy work without losing the relationship. By making the first response, the COI ping, and the bilingual content all run themselves, a 15-person firm can serve like a 50-person firm without sacrificing the personal touch the producers built their book on.
Bilingual market, manual back office?
We build AI comms layers for insurance brokerages serving multi-language markets. 30 minutes, free.
Meet with Hedgestone