transcribe

Integrate Cloud Ai Tools into your Existing AMS

Insurance Shoptalk 🎙️ · 39m · transcribed Jun 2026
More from Insurance Shoptalk 🎙️ Business
𝕏 Share ▶ YouTube 📥 PDF 🤖 .md

Transcript

0:00 a lot of the workflows that we we handle at this point come after the bind so say you've presented the customer this quote the customer says looks good let's do it they pay for it they sign the paperwork uh and you know the the carrier then presents a a policy document uh now Cara really steps into help with the servicing of that customer so if that customer has a question about their policy Cara has access to the policy

0:25 document and the policy information so Cara can read the question in your inbox and draft response for you welcome to insurance shop talk this is the podcast where the world of Commercial Insurance gets a little closer to home I'm your host Eric Stein and in every episode we're taking a deep dive into the insurance industry from the latest news and Trends to in-depth interviews with insure Tech experts carriers and agency newsmakers we're

0:55 covering it all so sit back tune in and let's start the conversation hi everybody today I'm excited to have uh nille console with me uh with oyster uh got a brand new product that involves a lot of uh AI That's combined in one system for agents to use for uh you know their agency management and really just integrating a lot of new technology into one cloud-based system uh nille thanks for coming on I really appreciate it

1:25 yeah it's great to meet you Eric thanks for having me absolutely absolutely and you said you are an in India right now and it looks beautiful uh is that a real background or fake no unfortunately it's it's not a real background uh yeah I am in India right now so it is uh past midnight my time but definitely getting used to the uh the times own difference here absolutely um as I mentioned I you we used some offshore folks in shenai and

1:51 it's been a 15year experience and phenomenal experience to be honest and I like uh luckily one of the gy jagen has been with us for literally 15 years so I get to kind of learn about the culture a little bit and all the monsoons and how how they you know they have to how do you guys how do they have to get to work it's just a crazy environment you know so depending on where you are

2:15 obviously yeah it's a it's a really diverse country you'll find all sorts of cultures all sorts of people languages food um all over the country right right bigger bigger uh bigger experience in America I know so let's kind of get into a little n nille why don't you kind of go into the new product the Cara uh that you launched at oyster um talk a little bit about your background as well and your role and um you know kind of set

2:42 the stage sure so I'll start by introducing myself uh my name is nikil I'm one of the co-founders at oyster I work primarily on product engineering and development so a lot of the AI features and the use cases that our customers come to us uh I work with our team our Stellar uh Team to put them all together into workflows that work for individual agencies so Cara is an AIC CSR that's geared towards automating sales and

3:13 servicing for insurance agencies so you can use Cara in in multiple different ways U primary primarily it's split into four different modules so there's a co-pilot that you can work with Cara to accomplish service requests uh and other operational tasks such as in forms certificate uh certificates of insurance uh checking for Eno exposure contract review and so forth there's a servicing module which helps agencies track open service requests assign them to Cara to

3:43 to work together and complete them there's a Communications module so our co-pilot and knowledge uh co-pilot and servicing modules can integrate with email and phone so however you talk to your customers Cara is there to assist and finally the base which is really everything that Karen knows and uses to adapt to your agency's workflows that's phenomenal yeah I mean and I've experienced using AI just you know separate components like chat GPT

4:14 you know separately so it sounds like you're trying to combine everything into one as much as you can what we're really trying to do is build something that works for insurance so I've spoken to quite a few insurance agents um you know everyone kind of r ranges on the spectrum of I've never even heard of AI to I use chat GPT every single day and there's unique problems that each of these personas face um I think we kind of have an intuition for

4:42 the potential impact of AI and insurance workflows but tools like chat GPT uh they're great for getting Quick Knowledge and and general um checks and and so forth um but I'm not sure how many people actually trust it to Output very accurate information especially in an industry like Insurance where accuracy and and risk is really really important so what we're building is the ability to take something like chat GPT and make it very very tailored to

5:10 insurance workflows and make sure that it's accurate and reliable as required by this industry awesome yeah I know the big push is always to integrate as many features into let's say a CRM or agency management system as possible and with just a big boys with you know veror and epic and uh you know all the all the different ones um it's it's always a challenge because you got a legacy system that you're used to working in

5:41 and you know have all your data in and your team's used to working in so it's hard to you know just change that if you will so would you would you also say it's an agency management system too or is it more CRM oriented I would say it's kind of adjacent to to both so some agencies find success in using Cara as a system of record uh usually those are smaller agencies that you manage things normally

6:09 through a spreadsheet or a to-do list um but for the vast majority of agencies that have an AMS Cara is not an AMS or a CRM but Cara works with your existing AMS and CRM to provide that intelligence layer on top so if there's there common workflows that you have to do you can use Cara to do them and Cara can re didn't write data to your AMS and CRM to accomplish those tasks like as I

6:34 mentioned we have an email integration so if a customers emailing you to uh get a certificate of insurance or hey does my policy uh meet these contractual requirements from my vender then Cara's in your inbox uh Cara is connected to your AMS Cara has the customer context and Cara can draft a response for you to review and send back to the customer awesome is that is that go through Outlook too or just the agency CR I mean SL that can

7:02 go yeah the email integration is agnostic to the client that you use so it can work with Google workpace Microsoft Outlook um any kind of email system that you use great okay and um would you say even Clos sort of systems like veror you know AMS 360 things like that you can still integrate with that so currently we have API Integrations with two ams's um that's hawksoft and epic uh and we're working

7:32 on getting easy links and AMS 360 uh in the next couple of months got it yeah so the these Integrations allow agencies to continue to work in a in a workspace that they're familiar with the agency Management systems and not need to maintain two separate systems of record upload and input information in two different uh you know workspaces we use these Integrations to make car work with your existing Tech stock that's

8:03 great yeah like I said people they hate the up upheaval of changing the entire system but you know they need these components you describe you know for AI um and and just for Automation and consolidation of vendors you know um and then kind of walk through like some I know you have like different components you mentioned and I on your website walk through a few that are the day in the life let's say have a CSR working on an

8:31 let's say a producer has a submission uh that they're working on and they're working on their apps uh and so it starts there and then it's gonna once he completes that it's going to be uh now given to his CSR to start marketing the account like walk walk through a couple of like real life stories of how that workflow you could assist with that workflow with Cara sure there's there's quite a few workflows that Cara can assist with uh I

9:02 think I'll talk about the top three that we've seen from our agencies um the biggest one by far is cross form fill so when you have a customer that is sending you information about their business you they're applying for insurance maybe you've had a phone call maybe there's some emails maybe there's some documents that go back and forth maybe there's an application of some sort uh you need to submit all this information in a structured form to a carrier or a

9:26 wholesale broker Accord form is typically to mind or Carrier specific broker specific supplementals uh what Cara can actually do is take the unstructured information that you have in phone calls emails documents disparate applications consolidate them all and generate the Accord 125 126 130 140 the carrier specific supplementals so that you're all ready to go without having to parse that information yourself fill out the forms yourself you're kind of relying on Cara to do most of the grunt work and then you'll

9:58 do a quick review before shipping it off so that's one of the most uh most requested use cases that Cara uh helps agencies with um I've got a couple more uh but you know I'm happy to take a pause here yeah yeah so dig into that so so let's say I got the Accord I started with some supplemental um is that really the basis of how carara knows what data to take and then try to transpose as much as

10:26 possible onto another let's say carrier supplemental exactly so Cara can read anything any kind of unstructured input and Cara can output documents based on the input so you can go from a chord form to a supplemental or an application to an accord form or an accord form to another Accord form so car is really good at reading unstructured information making it structured and then putting it into the output that you need uh in this case it's a document did it even go out to

10:58 the without being has even to go out to the client's website and Linkedin or what whatever public information and grab it kind like a chat GPT and and assist in even putting that information in yeah so there is there something that we're actually working on I'm glad you touched on it uh what we call it the underwriting module where as a part of this platform um you know as we expand Cara's capabilities you can ask car to

11:24 perform independent research and incorporate that information into the cord forms so for example based on the website and the Google search Cara could suggest a couple of GL sick or nakes codes for you to use on the application of course you know these are kind of information it's really up to the agent's judgment so we don't really uh we're not designing Cara to make decisions unilaterally rather Cara is more of a collaborative teammate where you have as the agent the final uh stay

11:56 on what goes into these forms and and what goes out of your agency got it all right so then do as much as you can with the client and the information you know to get to get Cara hopefully can help fill that in and then also complete multiple forms multiple submissions um and then I guess you could structure to even send the emails out once you approve it yeah so Cara does have an email drafting feature as well so what again

12:24 it's part of this workflow you can kind of construct it however you want if you if you want car to do independent research you just need to ask her if you want her to generate specific forms you ask her if you want her to draft an email to let's say a couple of carriers a broker maybe even the customer that is a part of that same thread that same workflow Carol will have that email ready for you to go and you can press

12:44 send and it'll Carol will send that email out got it and then I guess go into them more now let's say it's now it's been sent to multiple markets the proposals and quotes start coming back in and now you know you have to kind of compare and also look at policy data to make sure it meets the needs of the client for coverage right talk a little bit about that yeah yeah so that's actually the next use case that I wanted

13:12 to talk about Cara also has a coverage quote and contract comparison feature where you can upload kind of any combination of existing policies or maybe you want to compare um a customer's policy that's going to expire to a renewal policy or two or three quotes that you've gotten from multiple carriers uh or two quotes versus a contract that a customer has sent you and has asked which of these quotes is better for uh the insurance requirements

13:40 that my vendor is is asking for Cara can take any combination of those documents and give you a summary of the differences and a recommendation on what you should suggest to the client in terms of coverage once again you can you can incorporate other features of Cara in here so if you want Cara to then draft the email and send it to the customer you just have to ask but as those quotes start coming in from

14:03 carriers Cara is more than capable of comparing and contrasting them and outputting a summary and a recommendation awesome so then I've been seeing that the big concern is that even if you use chat GPT and they say they don't you know do this the uh data now is kind of going out there for all the AI to continue learning on in the in the in the world if you will how do you address the security of that client data

14:31 and proprietary you know information if you will so that all the other agents who may be using Cara and or other AI modules you know don't get access to that same information yeah it's a really uh complex question so there's a couple of different areas to break down number one when we handle customer data so if you've uploaded a quote contractor policy then each customer's data is segmented so no other agency has access

15:04 to any other agency's data uh either explicitly or implicitly Cara can only reference data for your agency uh and and no other agency which means that your data is not referenceable by other agencies uh and the second part of your question touched on how do we keep the data from being used in training so we use uh foundational models um which means that we have not uh done the you know 700 billion parameter training of a

15:33 model because that's really expensive uh but the way we use that model we use it in a virtual private Cloud so we have our own private deployment of that module which means of that model sorry which means any usage of that model doesn't feed data back to the parent company which means that data the customer data that we have is not able to be used for further training by foundational model providers the only only training for that agency uh and and

16:00 their proprietary you know um walled off area of Cara exactly and and that area of training is something that we are responsible for and not the foundational mobel provider uh so Cara can you know kind of as you use Cara more you you've uploaded more documents you've given Cara more feedback uh that's stored in a agency specific database that Cara can then pull from to learn about your workflows and the way that you you do business for example maybe you prefer to

16:30 write emails a certain way or maybe you prefer to see certain information in a quote comparison as you give car that feedback that gets stored in your agency's specific database and Cara then references that information thereafter yeah we create the r every time you want to do a request and like what is is are you is a proprietary or is it open AI I mean what is the backbone yeah so we use a combination of models so one of the models that we use

16:59 uh more often than the others is anthropics quad Sonet um it's a it's a pretty good module for comprehension and writing which are the two key areas that um you know insurance agencies really need to excel in right reading things correctly paying attention to detail and making sure anything that you write is correct clear and concise right sweet um so then moving forward so now you got the quote comparison you use that data

17:30 that you the car provides make the decision of which quote you want to present um now now do you want to talk about any other like workflow in that piece now what's the next steps sure so a lot of the workflows that we we handle at this point come after the bind so say you've presented the customer this quote the customer says looks good let's do it they pay for it they sign the paperwork uh and you

17:55 know the the carrier then presents a a policy document uh now Cara really steps into help with the servicing of that customer so if that customer has a question about their policy Cara has access to the policy document and the policy information so Cara can read the question in your inbox and draft a response for you referencing specific sections of the policy or whatever other documents are associated with the account and you can give it a thumbs up

18:22 and send it off or make some edits and then send it yourself um if a customer needs a Certificate of Insurance it's the same deal so if they need specific endorsements to be listed or if they need an additional insured um Cara can check the policy language and check all the endorsements um and make sure that what the customer is asking is provided for by the policy and if not let the agent know that hey in order to issue

18:47 the COI you might need to make these changes to the policy and when you say it's like sitting in your inbox like does it literally just email me as the agent or or c s r first so so to speak and then I get to for review and forward so you as the agent would receive or the CSR would receive the the request from the customer and then all you would have to do is forward that to

19:13 Cara so you can think of Cara as like an AIC CSR Cara will look at the the customer's account the request all the information and figure out what she needs to do and reply back to you the agent with whatever information uh or whatever partial completion of that request whether that's hey I generated the COI successfully because there were no issues please review and send it off or I couldn't generate the COI because this endorsement was missing or I

19:39 couldn't confirm that this language is valid in the description of operations got it got it okay cool hey everybody it's Eric I just want to pop in for a quick announcement exciting news I just launched my new book on Amazon it's called The Blueprint an insurance agency's unconventional rise to the top and it goes through a little bit about how I built the agency Insurance Solutions since 2009 to today and the blueprint terms of how I

20:07 actually did it with the tech stack the offshore Nearshore talent and digital marketing techniques that I use so pick up a copy today on Amazon thanks guys and then so keep going what else like on this this account so it can help with the service side uh you could it actually I'm assuming can it go ahead and issue the COI and you know send it to you to review exactly so it can create that Accord 25 form if uh based on the

20:37 information that's got from the customer and the information from the policy if it's got all the information then it can generate the the form for you for you to review and send out if certain information is missing for example uh maybe the the customer or the client forgot to mention who the certificate holder is or their address and Cara can say hey I've got most of the information but I'm missing this this in this piece could you please clarify with the

21:00 customer and let me know and I can finish the COI so Cara can identify missing information as well before generating the COI so then I I wouldn't need even my certificate system anymore I could just use Cara in general or do I still is it using my certificate system you could you could just use Cara um if your AMS if we have an integration with your agency management system and your AMS allows document upload then we

21:28 can push that information back into your AMS if that's where you would like the certificates to live but uh we also keep track of customer interactions and and all the documents associated with an account so we have a record of all the cois that have been issued for a customer through K okay so then yeah so that that's usually the reason you keep the system would be for you know prior information so um is it structured so that you know

21:54 at renewal I could just resend hit with a button say send all CER C ific holders from last year send the newer to them with this policy info and it would know to do that kind of like a search system yeah you would basically as part of your renewal workflow that would be something that you can let uh or at least configure Cara to do automatically like look at the certificate look at the renewal policy make sure that there

22:19 aren't any important differences like was an endorsement dropped uh if so that certificate is going to have to either change or we're going to have to make a change to the policy so car can can do that kind of review and reissue certificates now how technical do I have to be or you know to do this renewal configuration within Cara is that something you guys do do I get trained my team gets trained to do it how does

22:41 that work that's something that we're actively working on uh it's still a a very early feature that we're co-developing with our agencies but I essentially it's like designing a workflow so if you've used um you know workflows in HubSpot or Salesforce they're kind of this like drag and drop interface where you can say this is what I want to happen in this sequence for uh renewals you know I want the customer to be notified that they have a renewal

23:08 coming up uh I want this form to be generated from last year's application uh once a policy is available from the from the carrier I want this check in these the COI generation to run so you can kind of drag and drop to configure exactly what you want to happen here um the AI kind of runs in the background it's not really for Forefront feature of the product we're really focused more on uh core workflows that agencies have and

23:34 making them as efficient as possible makes sense can it also just like once you mentioned once an account's ready to bind can it go ahead and upload all the documents save it in the C the AMS and the right fields and is it trained to do that yet like an agent if the am agent if the AMS API allows us to do it then we certainly can do it so for for the uh for example epic allows us to

24:04 create and update policy records and account records uh via the API so if you're an epic user and you have API access you would just grant Cara that access and then Cara can update records and add activities and logs for example put customer calls in or emails in um that can also be put into your AMS but that really depends on the capabilities of the AMS uh so am you'd have to configure that say all Accords store

24:33 here and the notes about the client you know or client call can it if if allowed then can it document you know what was discussed on the call and can it actually listen to your call from the phone system and that s those sort things and translate it so so as part of the communication module that I mentioned Cara can actually take phone calls so you can decide if you want to forward all calls to Cara as the first L

25:00 of defense or set up Cara as an alternative to voicemail or just take calls after business hours and either in either way Cara provides this very realistic voice AI that allows her to understand what the customer requesting and kind of triage and handle that so for example if they're looking for uh a new application or a new quote Cara can uh prioritize that call and send it to an agent if an agent's available and If not then she can ask the questions that

25:30 an agent would ask to get all that information uh transcribed in a structured form for the agent to then see when they're available uh same thing for endorsements uh first notice of losses uh requesting certificates of insurance for things that require an agent the agent's tried first but otherwise Cara is perfectly capable of handling the initial collection and triage uh of information wrong you think it is it s I've seen some recently you know uh used by businesses that are

26:00 really good uh you know low latency very conversational you know like your chat gbd talking mode or Claude or whatever and you know Gem and I um how do you feel about the conversational and the realistic uh you know response and correct response latency T you know H how's that going yeah I I think the best judge of that are real customers um and I think you know what we what we've noticed is if you tell somebody who's evaluating

26:32 software that something is AI they pay very very close attention but as I mentioned we've been running the system at our uh Insurance Agency uh for some time now and I think for the last few months that's how long it's been deployed uh Cara is taking all of our phone calls um over 70% of customers can't even tell they're talking to an AI uh and the ones that can most of them are very impressed of course there are some customers um

27:01 who feel very strongly about talking to an AI um and for those customers you know they always have the option of saying hey connect me with a person and Carol will automatically say of course let me get you to an agent and an agent will pick up the phone but the vast majority of customers either can't tell or very impressed that they're talking to an AI well that's great yeah I guess again it's kind of nice that there versus

27:23 outbound which can be challenging um inbound at least there's certain you know a set of areas to focus on that you could get to get really good at you know about how the COI questions is an example you know yeah um versus outbound I tested a few things you know for or outbound and it's been very challenging because it's just too diverse and sounds too automated and I don't know yeah I think the natural sound and

27:56 the the flow of the conversation is a lot better better with some of the more Cutting Edge AI models especially the way that these text models and these speech models work together the latency is low um the the voice sounds realistic so there there will be like us and ums mhms like those kind of uh interjections and uh you know car is also very patient uh some customers need more help than other customers or or more um time to to

28:25 work through a request uh Cara has all the time in the world um so you can let Cara handle more in-depth uh customer requests as well and that could be that's you could have 20 calls or how many how many lines you have you could have 10 calls or be 10 caras going on at the same time even right exactly so let's say you're busy with a customer right now you still got Cara as the first line of defense taking

28:51 that phone call uh she'll transfer you as a as a customer if she can but otherwise she'll collect all the information note it down for the agent as a service request in the servicing module and let the customer know that the agent will call them back uh and that lets the agent kind of free up their time and focus on one of the most important most pressing customers at any given time and then schedule in the rest

29:14 of the customer service and support as their schedule allows and can Cara schedule appointment on cly or what have you as well yeah so Cara has uh a basic calendar we're working to expand this to more robust formats like for example if you have your outlook if you use outlook for your agency this isn't available just yet um but Cara could look at your availability and look at the urgency of the customer's request and find a slot

29:43 on your calendar that you can then move around uh to to call the customer back that's really cool yeah um and what else what else what other features do you think are be you'd like to highlight on the workflow we really like to work with our agencies um every agency's workflows are you know either completely different or slightly different and every customer that we work with we learn a little bit more about a new dimension of what

30:13 causes them pain so the way we like to work with agencies is we uh we'll set up an introduction called you know mostly to just discover and understand where are the inefficiencies what are the pain points if you had a magic wand uh and you could you know make something go away what would it be and then we'll tailor the solution to your agency so with the capability of AI and the foundational um intelligence that we've

30:40 built into Cara really any kind of insurance workflow is is possible that involves working together with with car so you can really think of it as an aicr that's constantly learning and building out your agency specific workflows so contract review quot and policy comparison um cois I've kind of touched on that another thing that Cara can do is review uh documents and outputs for Eno exposure so for example if you have

31:11 generated a Certificate of Insurance um sometimes uh mistakes can be made and that can expose your agency to Eno risk uh what Cara can do is is be a second set of eyes on that COI and instead of being the one to generate the COI Cara can look at the COI compare to the policy and say okay are all these fields actually correct are there any caveats is something checked on the co that shouldn't be checked does something not

31:38 apply to the specific certificate holder uh and give you a couple of bullet points and say like hey this this and this looks right although you might want to double check this this so that's another use case that we've co-developed with one of our customers basically reducing the um likelihood of Eno exposure through um out puts that aren't correct from an insurance perspective yeah really great yeah I can see how that could be helpful and then

32:06 as far as the pricing um how does that work is it like a SAS fee monthly fee or user licensing how do you set it up yeah it's it's fairly simple it's a per user per month fee so if you've got uh our standard pricing is um at the moment for for our initial set of customers it's $100 a month per user so if you've got one User it's $100 a month 10 is $1,000 a month uh and of course

32:35 you know uh if you commit to using Cara annually then there's some uh special benefits for that as well okay makes sense um yeah that's helpful versus you know trying to a develop this yourself or have Enterprise open AI Enterprise chat GPT trying to integrate that into what you're doing a full-time job you know yeah and car's really what we've heard about open AI from other insurance agencies is that they it struggles to

33:06 read policy documents in a way that um is it makes sense for insurance so some policy documents you know they're out of order uh sections might contradict each other and you need to understand how this carrier works and how these documents work to know exactly you know what's in effect what's covered what's not covered what the terms are that's something that chat GPT uh can sometimes struggle with because it is probably reading these documents linearly uh and

33:32 not considering the insurance specific meaning of the text um not to mention that there's limitations with the tool like chat GPT you're unlikely to be able to upload a 300 page policy document and have chat GPT read all of it whereas Cara is really optimized for these workflows Cara does really well with long long long documents and contradictory Clauses and endorsements that modify other endorsements and things being out of order so we've really kind of built that

34:00 foundational intelligence into carot to make all the workflows better yeah really is good to have an uh you know insurance industry industry specific you know um just like you're saying like I tried to type some things in on a workers comp class just real quick you know and it can sometimes can just uh you know not understand what what direction I'm going in because it's not Insurance you know trained like you're doing and you set it

34:26 up exactly [Music] PNC AIS have you run into bber which is a brooksher halfway company quoting workers comp for either your prospects or clients well great news insur Solutions a national mgaa has an appointment that is set up to allow you to quote bind and service byber at your agency and the coverage is in all states it's a-rated start are even accepted

34:59 there's 125,000 maximum premium and some of the tough Industries they write are construction transportation that includes Towing and parcel delivery manufacturing security guards that are armed and unarmed motor mediation and many more so go to our website at insureds solutions.net click the get appointed link we'll get your logins for you and your team and can set up a webinar for training if you'd like also feel free to email info insured solutions.net and I can uh handle the

35:31 submission myself or call 770 40253 43 be glad to speak to you and give you more detail [Music] thanks um well awesome well I mean this sounds pretty exciting I'm glad you guys are on the Forefront of getting this uh developed and working with some of these uh beta agencies to get it going and work out a lot of the Kinks and sound like you're you know pretty far along on it um if there somebody wanted to get in

36:03 touch you know with you guys to get some more info or get a demo what's the best way that they could reach out you know to you or your team yeah the best way would probably to be to go to our website get cara. uh there's a button there that you can click to to book a demo it's going to connect you directly with one of either me or my co-founder Vic um you know we talk to all of our customers uh

36:27 personally and directly just to make sure that what we're building and we're offering them it's going to deliver value immediately that's fantastic I'm glad you guys are intimately involved at this point you know to keep it on track um some exciting stuff going on in this world man uh this AI world so you guys are making it happen like real time in Insurance you know and making it simple uh which uh yeah again I've done some little more some research recently

36:56 like that outbound test I did you know it's uh it's challenging you know to figure it out yeah yeah there's a there's a lot of potential with AI and insurance I think the key challenge is getting it to work at the level that's required in insurance and there's definitely uh types of AI that you can build you know where if it works 70% 80% that's like good enough um but that's not really the case in Insurance like it

37:22 it needs to really nail the use case it really needs to understand Insurance to be useful can't afford to have something that's only 90% accurate in insurance so we're really focused on making sure that for any agency that we work with the accuracy is really high uh the confidence that you have as an agent in the software is really high yeah agreed agreed yeah it's and then you know as an agency owner you get very impatient too you know you want

37:49 everything to it's hard enough to get the team to start utilizing the the tools and if they don't work we lose confidence and then they don't get the process just don't get followed and so that you guys are trying to make it you know as perfect as possible which can be done because of AI you know can get to that 95% level versus some other you know software you know is a is a kind of a feed into

38:17 itself yeah it's definitely challenging but I think these are these are the problems that need to be solved right we don't we don't need another chap up we need to actually drive efficiency and value for for people uh and especially like agents and csrs and insurance I'm sure you know over the last you know 30 40 years a lot of tech waves have like come and gone uh and some of them insurance has gone on um relatively you

38:42 know after other uh sectors have but others have been kind of missed and I think AI is another one of those waves um that's different this time because of the way it deals with unstructured data makes it a perfect fit for insurance agreed agreed yeah I'm excited about this man I'm excited about where it's going to go and you know your progress so I want to have you on again probably in I don't know three or four

39:06 months I bet it'll be all different you know more even more robust yeah we're definitely moving very quickly uh so it's it's a really exciting time to to be working in AI especially in Insurance awesome well N I really appreciate you being on in your time and uh looking forward to uh hearing about the success and of course working together with you and of course thank you Eric for having me it was it was a pleasure to to talk

39:32 about Carol awesome man thanks again for tuning in to insurance shop talk if you enjoyed our Deep dive into the latest trends please don't forget to like subscribe and hit the notification Bell so you never miss an episode until next time I'm Eric Stein and this has been insurance shop talk [Music]

Summary

Oyster has launched Cara, an AI-driven customer service representative designed to streamline workflows for insurance agencies. Cara assists with various tasks, from drafting responses to customer inquiries to generating and comparing insurance documents, all while integrating seamlessly with existing agency management systems.

- Cara automates post-bind workflows, assisting agents with customer service inquiries and policy management.
- The platform features modules for service requests, communications, and document generation, enhancing operational efficiency.
- Cara can read unstructured data from emails and documents to generate structured forms like Accord forms and Certificates of Insurance.
- The AI can compare quotes and policies, providing summaries and recommendations based on client needs.
- Security measures ensure that customer data remains confidential and is not used for training other AI models.
- The system integrates with popular agency management systems and email platforms, allowing for smooth data flow and communication.
- Pricing is structured as a per-user monthly fee, making it accessible for agencies of various sizes.
- Continuous feedback from users helps tailor Cara's capabilities to meet specific agency workflows and needs.
© transcribe · For agents Built with care and craft by Gokul Rajaram