Sales AI agent
Quotes and guidance over WhatsAppWorkflow publishedHelps prospects, gathers the data needed to quote, and guides them toward issuing and paying for their policy.
For Carriers, Brokers, MGAs and Producers
A platform that can work hand in hand with your team to accelerate sales and automate your operation.
InsurMind
Shared core
Your producers and your team work on one shared core, and that core talks to the systems you already use.
Sales, and the operation that sustains it: retention, service and administration. Above them all, a layer that measures and oversees. Nine AI agents on one core.
AI agent · Sales
Quote, answer questions, capture data — including images —, offer discounts, compare competitors' offers and issue policies. Follow up on stalled opportunities.
AI agent · Invoicing
Generate and send policy invoices with the customer's tax details, under the carrier's rules.
AI agent · Reconciliation
Match payments received against issued receipts, identify discrepancies and keep the evidence for each case.
AI agent · Collections
Send payment reminders, provide payment links and check the status of receipts.
AI agent · Renewals
Identify expiring policies and prepare quotes to make renewal easy.
AI agent · Recovery
Run campaigns to win back lapsed or cancelled policies, picking up the customer data already on file.
AI agent · Service
Answer questions about the policy, send documents and handle change requests. Point to the carrier's channel in the event of a claim.
AI agent · Analytics
Measure campaign spend, conversion and the sales funnel to identify which channels generate business.
AI agent · Oversight
Review conversations, exceptions and evidence to spot incidents and make it easy for the team to step in.
From a single view, your team understands what each AI agent is aiming at, what it is executing and which cases need intervention.
Demo view of the InsurMind control center.
This is the pulse of your business. Spot opportunities, follow your AI agents, and decide where to act.
See their mission, operational activity, and where they need attention.
Helps prospects, gathers the data needed to quote, and guides them toward issuing and paying for their policy.
Reconnects with prospects in recovery campaigns to reopen the conversation and bring them closer to a new quote.
Operational indicators from the observed snapshot; these are not real-time data.
168 opportunities awaiting issuance. Review prospects who have already shown intent to buy.
Views by landing page, owner, and channel, with the source visible in the dashboard.
A view that helps your team follow the sales journey and decide where to step in: follow-up rules, payment reminders, discounts and campaigns. The screen below is a proposal with fictional data, not a live dashboard.
Percentages show progress between stages in a fictional cohort; cards show fictional cases in their current stage.
Scroll horizontally to see all six stages →Inquiry started; quote pending.
Asks about coverage options.
Has received a proposal.
Asks about coverage.
Replied; more details needed.
Wants to review the options.
Gathering details.
Documents to validate.
Awaiting payment.
Awaiting confirmation.
Payment confirmed in this example.
After-sales follow-up.
Mockup of the platform Kanban report. Fictional percentages, cards, and conversations. The pills indicate progress between stages; each card appears once in its current stage.
It quotes, captures the data, answers the questions and issues the policy without leaving the conversation. On the right, what the customer sees at each stage.
Sales We quote policies
The quote, delivered in the conversation.
Data capture for issuing.
It answers with the figure from that quote, not a generic one.
Reading the vehicle registration and issuing.
The AI agent leads the conversation. At any point, someone on your team can take it over to push the sale or resolve what only a person can resolve, and hand it back to the AI agent.
Adela steps into the thread that already exists. She does not start a new conversation or ask the customer to repeat anything.
From the same panel, with the suggested messages at hand. The customer sees her name, not a different account.
The conversation goes back to the AI agent, which picks up where the person left off.
Hable con uds por telefono para comentarles mi caso en particular sobre la renovación de mi póliza.12:02
No reply from the team yet.
Adela took the conversation
Adela · RenovacionesHola Juan Carlos, soy Adela del equipo especial de renovaciones. Voy a revisar tu caso y regreso a la brevedad.12:03
Adela · RenovacionesListo Juan Carlos, tu renovación quedó desbloqueada y ya puedes disfrutar de tu poliza con tranquilidad.12:07
Adela released the conversation. The AI agent picks the thread back up
Muchas gracias de verdad fue de mucha ayuda!12:08
Agente IADe qué Juan Carlos estamos para servirte.12:08
Take this conversation in order to reply.
Type your message… (“/” for quick replies)
Send messageInterest-free months reminder¡Hola! 🙂 Solo para recordarte: tu cotización la puedes pagar a meses sin…
Preferential price + ask for name¡Hola! 🙂 Solo para recordarte: tu cotización hoy tiene un precio preferencial si l…
General check-in¡Hola! 🙂 ¿que te parece la cotización? ¿Quieres conocer otras opciones o modalidades…
The receipt has a date, so the reminder has one too. The AI agent writes before it falls due, delivers the payment link in the same conversation and stops pressing as soon as the receipt is paid.
Each milestone is switched on or off separately, and they all go out at nine in the morning. If the customer asks to stop receiving them, they stop.
A policy that expires with no one warning is a policy lost. The AI agent spots the expiry, prepares the quote with the data it already has and puts it in front of the customer before they are left without coverage.
The customer does not have to look anything up or give their details again: the conversation arrives with the renewal already done.
Customers who cancelled for non-payment or whose term ran out. The AI agent quotes them again with what you already know about them and writes to them, one by one, in a campaign your team launches and measures.
The campaign is started, paused and resumed from the platform. Each customer receives a single attempt.
How far it covers, when it expires, what deductible it carries, where the PDF is. These are the questions that fill your team's phone, and the AI agent answers them by reading that customer's policy, not a generic brochure.
When the question goes beyond that, the conversation passes to your team with the full context, and comes back to the AI agent once it is resolved.
Today the tax details are gathered in the sales conversation and travel with the issuing. The Invoicing AI agent closes the loop: it generates each policy's invoice with those details and delivers it through the same channel, under the carrier's rules.
This piece is yet to be built. What already works today is capturing the tax details during the sale.
The AI agent pulls the policies and their receipts, matches them against the payments that came in and leaves every installment with a status you can defend: paid, pending, or with the discrepancy flagged and its evidence.
| Policy | Installment | Charge | Status |
|---|---|---|---|
| POL-•••• | 1 of 12 | Oct 05 | Paid |
| POL-•••• | 2 of 12 | Nov 05 | Pending |
| POL-•••• | 3 of 12 | Oct 05 | Declined attempts only |
| POL-•••• | 1 of 4 | Sep 12 | Promise to pay |
| POL-•••• | 4 of 12 | Aug 05 | Needs review |
When a payment does not match any receipt, the case is flagged with its evidence instead of disappearing into a spreadsheet.
Counting leads is not enough. The AI agent matches what you spend against what ends up as a paid policy, channel by channel, so you know where to put the next peso.
The six stages are those of the platform's real funnel. At each one you can see how many drop off, which is what tells you where the problem is.
An inbox ordered by whoever has been waiting longest, the status of every service the operation depends on, and an alert when the AI agent stops replying. Your team steps in where it is needed, not where it happens to be looking.
An integration failing is not news. It failing and nobody noticing, is.
With a defined goal and clear limits, it understands each case, takes the next step, and asks for help when needed.
It knows the quote, the alternatives, the data available and the exact point where each customer left off.
It looks up information, gathers data, triggers validations and requests operations from the connected systems.
It escalates exceptions, sensitive requests and cases that need commercial judgment or human attention.
InsurMind combines AI conversation and integrations with the carrier so that every action is useful, traceable and controllable.
Let's start with one concrete journey. We design the AI agent, define its tools, set its limits and connect the actions it needs to execute.