Record Retrieval Operations: The AI-to-Human Handoff
See how Orvera AI agents run record retrieval outreach, hand complex cases to specialists with the full file, and audit every interaction.
Watch an Orvera AI voice agent call a custodian of records, check a request’s status, capture next steps, and log the outcome with live monitoring and QA.
Overview
This demo follows a live AI voice agent through a record retrieval status call, the kind of outbound follow-up that litigation support and record retrieval firms make to providers and custodians of records every day. Anubhav Mohanty, SVP Product & Operations, explains the use case and then lets the agent place the call, confirming the request, checking what is pending, and capturing the invoice and turnaround details needed for the next step.
Alongside the call, the video shows live monitoring, where teams can observe concurrent calls and take over at any time, followed by the summary, transcript, recording, and QA score. It also covers how the workflow runs from a simple CSV upload or through API integration, with a zero data retention option for health information. The video is designed for operations and contact center leaders evaluating AI voice agents for high-volume workflows.
Key takeaways
Built for record retrieval. The AI agent calls providers and custodians of records to check request status, prepayment, and turnaround.
IVR to the right department. Because the use case is outbound, the agent navigates hospital and custodian IVRs before speaking with a person.
Stays on objective. Even when the representative is unsure, the agent captures the pending physician signature, the invoice process, and a contact name.
Live monitoring at scale. Teams can watch transcripts across 100 to 150 concurrent calls and take over any call.
Standalone or integrated. Upload a CSV to start dialing, or integrate through the API to pull requests and push outcomes back.
Zero data retention. Recordings and transcripts can be pushed to the client’s own SFTP server and deleted from the platform.
Anubhav Mohanty 00:00:00.00 Hello everyone, this is Aby again. Thank you so much for joining me for this video where I actually let you experience our live AI voice agents that our enterprise customers are today using.
Anubhav Mohanty 00:00:14.70 Today, let's go to another use case. It is a very interesting use case. It is a record retrieval follow-up calls. this use case is being used by our litigation support firms in the U.S. this we actually cater to multiple record retrieval firms in the U.S. through this and we are actually doing POC with a couple as well so there are three parts to this particular use case I'm actually going through one part today and then we will take the other parts in other videos so one is you know just to
Anubhav Mohanty 00:00:51.64 These are usually medical request forms for retrieving medical records or any other kind of records in which we have sent this request to provider office or custodian of records. It can be a release of information companies like Ciox and typically Datavant and MRO or basically provider's office or hospitals and stuff like that. So, we kind of call the AI agent kind of calls these providers or custodian of records to kind of get a status update on the request that was sent and anything that requires any changes that are required on these particular calls.
Anubhav Mohanty 00:01:30.28 uh now there are three type of calls one call can be a first initial calls that we are going to experience today in which basically it calls the provider to check the status of the request that was sent if it was received if not do they require uh any prepayment and stuff like that and then there is a follow-up call that the request was already sent we have not received the records and you know uh discussions regarding that uh there's also a kind of call which is called as which we can call these custodians to kind of confirm some of their details before we send the request so we make sure that the request is actually sent to the right location so we confirm what is the mode that they expect what is if fax what's the fax number if mail what's the mailing address is there any prepayment required what is the turnaround time etc so we have all the information with us already before sending the request so the chances of rejection kind of
Anubhav Mohanty 00:02:30.28 And, you know, the turnaround is also kind of shrink. In this particular use case, turnaround plays a very important role. And that is why, while implementing the AI agent, we take very important care of making it as efficient as possible. So the records can also come in in the least amount of days. So, yeah. So I think I'll let you experience that. Again, for this use case, it again, because it is an outbound use case, it again is well.
Anubhav Mohanty 00:03:00.28 efficient enough to kind of navigate through any IVRs of any hospitals or any custodians. It knows which number to select or how to, during conversation, how to kind of reach the right department. Once it reaches the right department and then start talking to your human agent, and that is when it kind of asks all the relevant questions. So yeah, I think that's kind of a context so you understand it as I always do. So yeah, let's get started. Oh yeah, and one more I make a call, I'll also let you experience our live monitoring module. So in the live call, I'll also open up, you know, the monitoring window so you can actually experience, see how the live monitoring works and see the transcript over there as well.
Anubhav Mohanty 00:03:45.50 So let me just open my platform. And here's the live monitoring. Currently, there are no calls going.
Live call 00:03:53.03 Hi, this is Christy from CallBotics. I'm calling to check this as the legal request sent to a truth record. Could you please help me with that? Yes, absolutely. I can. Can you tell me the patient's name, please? Sure. The patient's name is David.
Live call 00:04:16.56 all right um can you spell the last name for me please
Live call 00:04:24.14 yeah it's spelled as m-a-r-t-a-n okay and what's the date of birth
Live call 00:04:33.32 yeah um it's July 23rd 1985
Live call 00:04:39.92 okay um let me try to find it Um.
Live call 00:04:47.00 Sure. Take your time. Uh. When was this sent? And how was it sent to us?
Live call 00:04:56.34 Yeah. It was sent by Rock. And it was sent on April 17th. 2025.
Live call 00:05:04.68 Okay.
Live call 00:05:06.88 Um. Hi. Let me know if you need any more info. Are you looking at that? Yeah. and what was the fax number that it was sent from?
Live call 00:05:23.94 Yeah, so the request was sent fax number 312-555-7890 and it was sent from the fax number 800-238-730-07.
Live call 00:05:40.92 okay and uh and what what kind of documents are you really looking for
Live call 00:05:51.68 yeah um we're looking for medical records for david marnon the date range we need is from january 1st 2025 to march 31st 2025 okay all right so yeah I see the request over here uh uh it's in progress I think I think we are pending a physician signature on this one
Live call 00:06:17.64 oh thanks for checking that so um some charges waiting on the physician signature Do you know by when that might be completed?
Live call 00:06:27.92 No, I'm not sure, but I think it should take a few days.
Live call 00:06:37.06 All right, that helps. So should I just call again in a few days? We haven't received anything yet. And just confirm, are the records handled in-house or do you need to copy service for these? We actually handle it in-house.
Live call 00:06:53.90 Oh, got it. Thanks. And is there any type of procurement invoice that we should expect or is there just any completion of the physician signature right now? Yeah, I think as soon as the physician signature is completed, we'll send you an invoice and I think you should receive it and you can pay it up and then we'll release the records.
Live call 00:07:19.14 All right, that's clear. Guess I have it right for a note. Once the doctor signs, you'll find an invoice, and after you pay, you'll release the records. Got it. And can I have your full name for future reference? Yeah, sure. My name is David Smith.
Live call 00:07:43.82 thank you david smith I appreciate your help today I'll wait to hear from you once I can truth on any center voice I hope you the good one goodbye goodbye take it
Anubhav Mohanty 00:07:59.12 all right so that was our call uh like you saw I kind of tried to uh ask different questions if you know it gets confused uh but it kind of stuck to its objective uh it also took some additional information like what's the turnaround time should they call back uh what's the fax number and stuff like that so uh now let me show you the outcome quickly uh and you also by the way saw the live monitoring window so this is this is a feature of our platform so any calls that's going you know your our enterprise customers team kind of can live monitor the calls like no it was just one call going but you know some of our higher volume use cases they have concurrency like 100 concurrency or 120 or 150 concurrency going so you can monitor all those calls over here we have a we have created it in a way that you can have multiple people on your team who can observe observe kind of live transcripts so and and whenever they want they can kind of disconnect
Anubhav Mohanty 00:09:04.68 the call or kind of take over the call so uh so yeah that's kind of an advantage so I I wanted you to kind of uh see though how that kind of uh looks and now let's come to the reporting section uh
Anubhav Mohanty 00:09:27.10 yeah so this is the call for for patient david martin the request was sent april 17th for records from january 1st to march 31st one sign and invoice will be sent and records will be released upon payment the records are handled in-house christy was advised to follow up in a few days if not received so uh the facility contact is david smith so and this is again the entire transcript and the recording and again the QA score call was trade I'll actually try to uh do a bad call so that you can see a bad QA score uh but let me let me try to do that uh but yeah Overall, this is how the output looks like. You've already seen the call logs that you kind of get with this.
Anubhav Mohanty 00:10:20.06 And typically, this can also kind of, and again, because this is an outbound use case, as I always say, it is very easy. This can run as a standalone platform without any integrations required with your existing system. it is as simple as you uploading a CSV file or excel file to a particular to this u or to this UI and the system starts dialing as soon as the dialing is done you can download the report and you can upload it into your system but if you have to do an integration then obviously that's always an option and we have we have done integrations with some of our most of our clients currently using this particular use case have their own proprietary softwares uh so we have actually integrated with their proprietary software very easy to integrate through our uh our API set uh we actually pull the requests uh in the start of the day they give us what dispositions you want to pull we pull all those requests the system that is the call
Anubhav Mohanty 00:11:21.98 and it again pushes all the outcomes as in when the call are done along with the recording so one of our current customers actually wanted the recordings not to be on our platform but the recordings to be on their platform at the SFTP server so we are kind of pushing all the transcript and recordings to their platform as well as soon as the calls are done and we kind of delete it from our platform so we by default kind of believe in a zero data retention policy so if you don't want us to retain anything we just want to process the call so that the outcome is there and you can outcome and insight and as soon as the outcome is generated we can kind of push all the outcomes to your server or your systems and we kind of delete it delete everything from our site including the logs so uh and that's the benefit of using it in health information management and healthcare kind of a use case uh and similarly like our litigation support uh clients this is most uh mostly to do with health information management so they are they kind of feel very secure handling with the
Anubhav Mohanty 00:12:27.84 So that's about today's use case and how we are working with record retrieval follow-up requests. Like I said, in one of my next videos, I'll also let you experience one of our more complex follow-up calls as well in which you kind of experience what if things go, things does not go well, what if things does not go right. So yeah, we'll do that in one of our calls in the future as well. for joining me I hope uh you kind of like the output uh for this particular use case as well and next time I'll again come with another exciting use case that we are currently working on till then thank you so much goodbye and good work
Keep watching
See how Orvera AI agents run record retrieval outreach, hand complex cases to specialists with the full file, and audit every interaction.
See how the Orvera dashboard powers a live AI voice agent handling open enrollment calls, eligibility checks, portal guidance, and real-time escalation.
See how Orvera AI connects with 500+ ERP, CRM, and CCaaS platforms to bring customer data, conversations, and workflows into one intelligence layer.
The same agentic voice, agent assist and quality workflows, running on your own call flows in 3 to 6 weeks.