AI for Insurance Agencies: Building Better Workflows with Cara
A prospective client calls your agency on Saturday night.
They want a quote. They have time to answer questions. Your team will not be back until Monday.
What happens next?
That question sits at the center of David Carothers’ conversation with Nikhil Kansal, Co-founder and CTO of Cara, on the Power Producers Podcast.
At Florida Risk Partners, David wanted a better way to handle incoming calls, collect information, and move opportunities into the agency’s sales process. Remote work had made phone coordination more difficult, while personal lines and smaller commercial accounts required a more efficient approach to intake and follow-up.
His starting point was specific: improve the phones.
The resulting workflow extended much further. Cara became part of a connected process involving HubSpot, quoting systems, document comparisons, proposal preparation, and automated communication.
In this episode, David Carothers and Nikhil Kansal explain how that process works, where people remain involved, and what agency owners can learn from it.
The conversation offers a practical look at AI for insurance agencies through an operating example. Its value comes from understanding how individual tasks fit together.
Watch the full conversation with David Carothers and Nikhil Kansal on YouTube.
Start With an Agency Problem You Can Define
David’s first concern was the amount of human attention required to handle incoming calls.
When a call arrived, multiple phones could ring. Someone needed to answer, determine the request, collect information, and direct the conversation.
Those interruptions were particularly difficult to justify when the agency was handling smaller accounts.
A producer might stop a higher-value activity to gather routine information. An account manager might have to switch away from technical work. After hours, the opportunity could sit until someone returned.
David wanted to make that initial interaction more consistent.
This is a useful starting point for an agency evaluating AI.
Identify the task, the people involved, and the effect on the business. What takes too much time? Where does information get lost? Which requests repeatedly interrupt the team?
A clear problem gives the technology purchase a purpose.
It also gives the agency something to evaluate after implementation. Answering calls more reliably is an observable change. So is collecting the information needed for the next step.
Understand the Insurance Work Behind the Product
Nikhil explains that Cara grew out of the founders’ experience running a digital insurance brokerage.
They encountered the same work familiar to agency teams: speaking with clients, collecting underwriting information, preparing submissions, reviewing documents, and coordinating sales and service.
They built technology to make those activities more efficient.
The brokerage was eventually sold, and the software became the basis for Cara.
That history helps explain the product’s development.
Voice intake was an early focus. From there, the team worked on moving collected information into ACORD forms and supplemental applications. They also developed a knowledge base to help agency employees navigate carrier information and placement questions.
The recurring goal was to reduce repetitive work while making insurance knowledge easier to use.
For an agency owner, the relevant question is how well a product understands the actual workflow.
Can it collect useful information? Will that information reach the right system? Can your team review and use the output without rebuilding it?
Those details determine whether automation creates a practical improvement.
Connect the Phone Conversation to the Sales Pipeline
David describes Cara answering incoming calls and identifying what the caller needs.
For a new quote request, the workflow distinguishes between personal and commercial insurance and collects relevant information.
At Florida Risk Partners, that information then enters HubSpot.
David explains that the configured process creates the contact, the company where appropriate, and the deal record. The opportunity is connected to the relevant pipeline.
That connection is central to the example.
The phone interaction produces something the agency can work with. The team does not have to begin by listening to a voicemail, copying details, creating records, and deciding where the request belongs.
The opportunity arrives with context.
This also makes after-hours intake more useful. A caller can provide information while the office is closed, and the agency can begin the next working day with a record in its existing process.
The workflow requires configuration. Routing, fields, and pipeline stages need to match the agency’s operations.
David’s example shows the value of building that connection deliberately.
Give Each System a Clear Job
One of the most important points in the episode is that Florida Risk Partners uses several systems together.
Cara contributes to the workflow. HubSpot manages pipeline activity and follow-up. Quoting systems and carriers perform other parts of the process. Agency employees review information and handle decisions that require their expertise.
Understanding those responsibilities prevents unrealistic expectations about what one product will do.
| Workflow stage | Role in David’s agency example |
|---|---|
| Initial phone conversation | Cara identifies the request and gathers information |
| Contact and opportunity records | The configured integration places information into HubSpot |
| Follow-up and reminders | HubSpot automations respond to pipeline stages and communication preferences |
| Rating and quoting | The agency uses its quoting systems and carrier workflows |
| Comparison and proposal preparation | Cara helps organize quote information and prepare documents |
| Review and client advice | Agency staff verify information, explain recommendations, and handle questions |
The efficiency comes from the handoffs.
If a caller provides information but the team still needs to re-enter it several times, much of the benefit disappears. If a deal record is created but nobody owns the next step, the opportunity can still stall.
Each system needs a defined contribution, and each transition needs to work.
Use Automation to Improve Small-Account Economics
David’s interest in personal lines and smaller commercial accounts created a particular operational challenge.
These accounts can involve substantial communication relative to the revenue generated.
Repeated phone calls, missing information, application preparation, and follow-up can consume the margin in the account.
David uses a pressure-washing niche as an example of a workflow his agency has streamlined.
In the process he describes, Cara collects intake information. The agency then uses a suitable carrier workflow to obtain a quote and sends the customer a link through its communication templates.
For suitable risks, the customer can review the quote and proceed through the carrier’s purchase process.
David reports that some of these opportunities can reach a bindable quote in under five minutes.
That timing belongs to the specific workflow he describes. The broader lesson is how a defined niche, a suitable market, and a repeatable intake process work together.
The agency understands the questions it needs answered and the next step it expects to take.
Automation then reduces the routine effort around that work.
This makes account selection and process design part of the technology decision. The agency should understand which risks fit the streamlined path and which require a more involved review.
Let Clients Choose How They Want to Communicate
David describes another change that improved the agency’s follow-up: asking prospects for their preferred communication method.
Previously, the agency could send a text and receive no response. The team had not established whether texting was how the person wanted to engage.
The revised process gives the prospect a choice.
The workflow then uses that preference to guide subsequent communication.
This is a small design decision with a significant implication. A message being delivered does not mean it is reaching the client in a useful way.
A customer may prefer email because they need to send documents. Another may respond more readily by text. Someone else may need a phone conversation.
At Florida Risk Partners, these preferences become part of the CRM process. David also describes texting being used where the person has opted in.
That allows automation to support a more responsive experience.
The agency is asking how to make the next interaction convenient, then using the answer.
Build Follow-Up Around What the Opportunity Needs
David explains that HubSpot automations are connected to pipeline stages.
When the agency is waiting for information, the opportunity remains in that stage and receives the appropriate reminders.
Once the required information arrives, the deal can move forward.
This gives follow-up a practical purpose.
The message should tell the prospect what is missing and how to provide it. The agency should also know what needs to happen before the next stage begins.
A pipeline therefore needs more than stage names.
For each stage, the team should understand:
What information is required.
Who owns the next action.
What communication should occur.
What allows the opportunity to advance.
When a person needs to intervene.
These definitions make automation more reliable.
They also reduce the burden on employees who would otherwise need to remember every incomplete request and prepare each reminder manually.
The process can remain active without requiring constant attention from a producer.
Show Clients the Reason Behind the Recommendation
David’s use of Cara extends from intake into quote presentation.
He describes a personal lines process in which the agency obtains quotes, reviews the relevant options, and uses Cara to prepare a side-by-side comparison.
The team then checks the information and the proposed recommendation.
That review is an important part of the workflow.
The comparison helps organize the options. Agency expertise determines whether the information is accurate and the recommendation makes sense for the client.
David explains why showing alternatives matters.
A client who receives only one proposal may wonder whether the agency considered other options. They may not understand what drove the recommendation or which trade-offs were involved.
A comparison gives the agency a way to explain its work.
A lower premium could come with a different deductible, sublimit, or settlement provision. An option costing more may include something the client values.
Those differences need clear language and verification against the actual quote documents.
The result is a more informed client conversation. The buyer can see why the agency recommends an option and what the alternatives would mean.
Combine the Recommendation With the Supporting Information
Florida Risk Partners also uses Cara’s proposal tools with a branded agency template.
David describes preparing the recommended proposal and combining it with the comparison material.
The client receives the proposed option along with information about the other choices considered.
This creates a useful structure:
The proposal explains what the agency recommends. The comparison explains the reasoning.
That structure can make the producer’s work more visible.
It also gives clients a reference when discussing the decision with someone else in the household or business.
The connection to David’s conversation with Guffy Wright about value and client alignment is clear: buyers need to understand how a recommendation relates to what matters to them.
Technology can help prepare the explanation. The agency still needs to know the client well enough to make it relevant.
Reduce Repeated Switching Between Tasks
David describes his team working through several opportunities in sequence.
While one quoting process runs, an employee can address another task. Once those results are ready and reviewed, a comparison can begin while the next account enters the quoting process.
Proposal preparation becomes another defined step.
He compares the rhythm to a factory.
The useful point is the reduction in repeated setup and interruption.
Employees do not have to begin every account from an unstructured collection of emails, calls, and notes. They can work through a consistent process with the information they need.
Nikhil connects this to giving agency staff more time for work where their expertise matters.
That might include discussing a complex quote, developing a relationship, or evaluating an account that requires greater attention.
Agency owners can look for the same opportunity in their own workflows.
Where does the team repeatedly switch tasks? Which parts of the process could run in the background? What information would make the next action easier?
Those questions can reveal useful changes before another tool is purchased.
Explore Document Review Without Removing Professional Judgment
David also discusses potential uses involving commercial policies, certificates, and subcontractor agreements.
His example concerns the difference between information shown on a certificate and the coverage provided by the actual policy and endorsements.
He sees value in using comparison tools to identify items that deserve closer examination.
That is a sensible way to frame the workflow.
The software can assist with organizing documents and highlighting possible discrepancies. A qualified person still needs to evaluate the wording, understand the request, and communicate the finding appropriately.
A polished comparison is a starting point for review.
The agency needs a defined scope for the work and a clear owner for the conclusion. If a matter requires clarification from a carrier or another professional, the workflow should provide for that next step.
David’s broader point is that preparation time affects which services an agency can offer consistently.
A tool that reduces routine document handling may create capacity for more useful analysis and client conversations.
The value depends on the review that follows.
Keep the Path to a Person Easy
David wanted to know what would happen when a caller preferred to speak with an employee.
That concern influenced the phone configuration.
He describes working with Cara and the agency’s phone provider to establish a dedicated transfer route back to Florida Risk Partners.
The implementation detail matters because the customer experience depends on the complete call.
An AI assistant may collect information effectively, but a caller with a time-sensitive service question still needs an appropriate route to the team.
The agency should test that route.
What happens when someone asks for an employee? What happens outside business hours? How does the request reach the right person? Does the team receive the context already collected?
David also describes Cara identifying itself as the agency’s AI assistant when callers ask.
A clear role and a usable transfer process help the customer understand the interaction.
The goal is to make contact with the agency easier and prepare the team for the conversation that follows.
Treat Implementation as Ongoing Operational Work
David shares a simple example of a routing problem.
The initial conversation did not always clearly establish whether the caller wanted personal or commercial insurance. Some leads therefore reached the wrong pipeline.
Adding that question earlier helped the system collect the right information and route the request appropriately.
The lesson is practical: testing reveals requirements that a demonstration may miss.
Agency teams need to examine real calls, records, and handoffs after launch.
Are the questions useful? Are the required fields being completed? Does the information reach the intended location? What happens when a caller gives an unexpected answer?
David describes an ongoing line of communication with the Cara team.
Nikhil emphasizes the work involved in building, testing, maintaining, and improving a reliable product.
That continuing work is part of what an agency evaluates in a technology partner.
The product needs to function today, and the relationship needs to support improvements as the agency learns.
Evaluate the Full Cost of Building It Yourself
David acknowledges that some insurance professionals enjoy developing their own tools.
He also questions whether that is the best use of his time.
His focus is writing business, operating the agency, and spending time with his family. Building and maintaining software would compete with those priorities.
Nikhil distinguishes between producing an early demonstration and maintaining a dependable system.
A prototype can show promise. A business workflow needs to handle changing inputs, integrations, corrections, and ongoing use.
For an agency owner, that means examining the whole commitment.
Who will maintain the tool? Who will test changes? What happens when an integration stops working? How much time will the team spend troubleshooting?
Those questions support a more complete comparison between internal development and a vendor solution.
The decision should reflect the agency’s capabilities and priorities.
A product purchase also requires effort, but it can place more of the development and maintenance work with a dedicated team.
Measure the Workflow’s Contribution
David reports that the agency’s use of Cara produced results he could connect to incoming opportunities and subsequent business.
He looked at the types of accounts, when the leads arrived, and how quickly the agency responded.
That measurement approach is more useful than judging the system by how impressive a demonstration looks.
An agency can assess:
Whether after-hours requests become usable opportunities.
How much information is complete at intake.
How quickly suitable prospects receive the next response.
How much manual follow-up the team performs.
How long comparison and proposal preparation take.
What happens to opportunities after they enter the pipeline.
The agency should also consider implementation time, software costs, and ongoing review work.
This creates a clearer picture of the contribution.
Results will depend on lead volume, account mix, existing processes, and how the workflow is configured. The goal is to understand what changed in the agency’s own operation.
Put People at the Center of the Improved Process
This episode complements David’s discussion with Ashley Napier about employee development.
Ashley emphasized clear expectations, supported practice, and coaching. Nikhil describes technology that can reduce repetitive work and help employees use their expertise more effectively.
Those ideas fit together.
A faster process still needs people who understand the work. Employees need to know how to use the tools, review the outputs, recognize exceptions, and communicate with clients.
Automation can create capacity. Leadership decides how that capacity is used.
If the team spends less time gathering basic information, the agency can direct more attention toward advice, relationships, and development.
That outcome requires an intentional operating plan.
Begin With One Connected Workflow
The episode offers a practical starting sequence for agency owners.
Choose a recurring problem. Identify a task with visible friction, such as intake or missing-information follow-up.
Map the current process. Show where information arrives, where it goes, and who acts on it.
Define each system’s responsibility. Make the roles of AI, the CRM, quoting tools, and employees explicit.
Configure the handoffs. Establish required fields, routing, pipeline stages, and the conditions for human involvement.
Test real scenarios. Include incomplete requests, after-hours calls, and callers who want a person.
Train the team. Explain the workflow and provide practice with both ordinary requests and exceptions.
Review the results. Measure time, information quality, client response, and the business generated.
Expand from evidence. Add another use case once the first process works consistently.
This gives the agency a manageable way to learn and improve.
Build an Agency That Makes Better Use of Its Expertise
David began looking for a phone solution.
The larger opportunity emerged when he connected intake to the agency’s CRM, communication, quoting, comparison, and proposal processes.
That is the central lesson of this episode.
AI becomes useful through the work it helps the agency complete. Information needs a destination. Opportunities need a next step. Recommendations need review and explanation. Clients need a clear way to reach the team.
When those pieces fit together, the agency can make better use of both its technology and its people.
For owners evaluating AI for insurance agencies, the next step is to identify a workflow worth improving and examine how a solution would operate inside it.
Watch the full conversation with David Carothers and Nikhil Kansal, connect with Nikhil on LinkedIn, and visit Cara’s website to explore the platform or request a demonstration. Find more conversations about sales, leadership, and agency operations in the Power Producers Podcast archive.

AI for Insurance Agencies: Building Better Workflows with Cara
A prospective client calls your agency on Saturday night. They want a quote. They have time to answer questions. Your team will not be back
Value-Based Insurance Sales: Stop Chasing Quotes and Start Creating Alignment
Guffy Wright once lost an opportunity despite presenting an insurance proposal that was $2 million less expensive. He believed his coverage was stronger, too. He

Insurance Agency Employee Development: Build Stronger People Before You Replace Them
An agency owner looks at a phone report and sees a problem. One employee’s calls average 30 minutes. That seems too long. The owner starts

Insurance Agency Culture: Building a Team Great Producers Want to Join
A producer walks into a sales meeting with good news. An introduction worked. A decision-maker agreed to meet. An opportunity is moving forward. Then the

Building a Data-Driven Insurance Agency: Premium Is for Show, Revenue Is the Dough
An agency can write more premium and still struggle to understand whether the business is becoming more profitable. Premium volume tells part of the story.

Insurance Intake Automation: How Better Data Creates Better Client Conversations
A prospect agrees to talk about insurance. Then the work begins: locating declarations pages, answering questions, sending documents, and repeating information that already exists somewhere
Responses