AI Structuring Data from Unstructured Medical Documents | Minds Medical #384
Executive Summary
In this episode, Joe interviews Lukas, co-founder of Minds Medical, and Ingo, senior partner of SYNPULSE. The story follows the pivot of Minds Medical from a healthcare-focused AI startup serving hospitals to its services for fintech/insurtech companies during the COVID-19 pandemic. They also discuss the challenges of extracting structured data from unstructured medical documents and the potential applications of this technology in the insurance industry. The conversation highlights the joint offering of intelligent document processing for the insurance industry and the acquisition process between the two companies.
When Corona hit, all hospitals stopped picking up the phone. So we had to focus all our energy on the insurance product.
Lukas Naab, Founder Minds Medical
Follow Us!
If you are listening to our content for the 2nd time please consider following us on social media.
With SYNPULSE8 we are now building a product for the global insurance market.
Lukas Naab, Founder Minds Medical
Our Enabler HTAI and the Enterprise Europe Network Hessen
This recording was made possible by HTAI and the Enterprise Europe Network Hessen. These organizations have made tremendous contributions to helping startup businesses succeed and thrive, providing a range of services from helping to find grants to ongoing partnerships. By taking advantage of these resources, startup companies can network and develop innovative strategies for success on the international stage. The dedicated support of HTAI and the Enterprise Europe Network Hessen is paramount in providing startup businesses with the tools for lasting success. Learn more here: https://www.htai.de/ and https://www.een-hessen.de/
Our Enabler Startupraven
A startup’s journey can be a tough one, but it doesn’t have ́to feel like you’re alone on your quest! Invest in others’ success with us by joining our community of entrepreneurs who are building amazing things every day — no matter how big or small their ideas may seem at first glance.
The best way to find investors and cooperation partners for early-stage startups. Sign up here:
We discovered for example that one large hospital in Germany has more than 200 different forms just to discharge patients. [And there are more than 1.800 different hospitals in Germany]. This amount of unstructured data requires machine learning and AI to be useful.
Lukas Naab, Founder Minds Medical
The Video Interview is set to go live on Thursday, May 25th, 2023
The Audio Interview
You can subscribe to our podcasts here.
The Founder
Lukas Naab (https://www.linkedin.com/in/lukasnaab/) is the founder and CEO of Frankfurt-based AI startup Minds Medical. Since the acquisition of Minds Medical he is in the new company the principal product owner for SYNPULSE8 (https://synpulse8.com/). At the outbreak of corona, hospitals had many other things to do and just stopped picking up the phone, all at the same time. So Minds Medical had to focus completely on its insurance product. They eventually decided to join forces with SYNPULSE since they already had a joint offering for disability insurance providers.
The Startup
Minds Medical was acquired by SYNPULSE, now known as SYNPULSE8 (https://synpulse8.com/). This becomes part of their offering to insurance companies and insurtech startups, active in the healthcare space. The problem these insurers are facing is the sheer amount of unstructured medical data they get bombarded with to make a simple medical decision.
SYNPULSE8 helps to structure this data and make it conveniently available for the insurance underwriters, who have to make the decision to insure the applicant or not.
July 2019 Interview
You can find the interview here:
Tune in to our Internet Radio Station here:
Feedback
We are always looking for ways to make the show better. Please take this opportunity and share your feedback with us! You can also suggest topics or interview partners that you’d like featured on future episodes by leaving a comment below — we would love to hear from YOU!!!
The Interviewer
This interview was conducted by Jörn “Joe” Menninger, startup scout, founder, and host of Startuprad.io. Reach out to him:
www.linkedin.com/in/joernmenninger/
https://twitter.com/JoeMenninger
Keep Up to Date
Here is our publication calendar: https://calendar.google.com/calendar/embed?src=012i27ak525qi3smgjl8v9tbio%40group.calendar.google.com&ctz=Europe%2FBerlin
Topics Discussed in this Interview
In this interview, we are talking about
#Innovation #startup #startups #startupradio #entrepreneurship #entrepreneur #business #motivation #success #entrepreneurship #mindset #goals #entrepreneurlife #lifestyle #businessowner #believe #positivevibes #branding #innovation #techstartups #germany #startupinvesting #startupnews #startupcompany
#MindsMedical #insurance #insurtech #health #healthcare #medtech #medtechstartup #aistartup #ai #ki #fintech #machinelearning #structureddata #unstructuredmedicaldocuments #insuranceindustry #underwriters #claims #electronicpatientfiles #intelligentdocument #dataprocessing #financialservices #consulting #technologycompany #documentprocessing #EENHessen
Automated Transcript
Narrator (00:00:05) — Welcome to Startuprad.io, your podcast and YouTube blog covering the German startup scene with news interviews and live events.
Joe (00:00:20) — Hello and welcome everybody. This is Joe from startup rate.io in our series, together with Hassan Tradein Invest and the brand European Enterprise Network. We present to you today and interview in our sub podcast Tech Startups, Germany. Therefore, I would like to welcome a returning guest. Hey, Lucas, how you doing?
Lukas (00:00:44) — I’m doing great. Hi, how are you?
Joe (00:00:47) — I’m doing good as well. We may hint to our audience that you are a returning guest, having had our last interview back in 2019 with the Startup Mind Medical. And apparently something has changed. Um, the thing that changed is Ingo here because, um, you guys got acquired and Ingo, can you introduce yourself?
Ingo (00:01:14) — Yes. Hello. My name is in, as you can see, I’m sitting in the car, not riding the car, parking it, and, uh, enjoying the podcast here. So I’m, um, a senior partner with, uh, SIM Pauls in, uh, a global tech and consulting company, and I will tell more about our story with, uh, Lucas and Mines medical later.
Joe (00:01:37) — Great. Lucas, can you first take us a little bit on the right, uh, that is known to everybody as Corona and recession? Um, how we got through with Minds Medical since let’s say Mm. Rule of thumb, uh, outbreak of Corona
Lukas (00:01:56) — . So, um, as, as some of the listeners might remember, mine’s medical, um, is a machine learning company structuring unstructured data in the healthcare industry and insurance industry, and, um, focus on hospitals back then still in 2020 to automate processes and, uh, reimbursement processes in hospitals. And when Corona hit, um, basically all the hospitals star, uh, stopped picking up the phone because they were obviously quite busy with, uh, doing other things. And, um, we sort of had to figure out how to, how to deal with that and, um, simples and that we managed to, um, work through that was no part, was no small, small part, uh, thanks to Simples and, um, and the great people over there because we just had started a corporation in the insurance industry for, uh, uh, risk, uh, pre inquiries in the, in the disability insurance market, and to build a joint program or a joint, uh, joint application there together.
Lukas (00:03:05) — And so we completely shifted our focus from only healthcare or mo mostly healthcare, uh, to, to health insurance and life insurance. And obviously it was a tough time. We, back then, we were, our 10, 10 people in the company had just, uh, gotten another, uh, funding and, and kind of the things that also helped, uh, , uh, surviving, surviving Corona, so to speak. But, um, we, we pivoted, uh, and basically left the hospitals, um, to be hospitals, and do their thing and, uh, refocused completely on the insurance industry. And, um, yeah, got to know simples a lot better. We knew each other since 2017, I believe. But, uh, started the joint venture there together and approaching clients and building a great product. And, um, yeah, that’s, that’s one of the reasons we survived, even though with a little smaller team than,
Joe (00:04:03) — Well, main goal is survival. So congratulations on that. Um, when you’ve been talking, you could either classify yourself steadily as a big data, as an AI startup, as an analytics startups, or you could say, well, we’ve been a MedTech startup and now we are pivoted to be a FinTech or in Intertech ,
Lukas (00:04:23) — We are ai, uh, structuring unstructured data in FinTech and intertech. So we are joining, we are, we have a, we have a clear, uh, focus industry focus, uh, but also clear technology focus for, uh, intelligent document processing, basically.
Joe (00:04:42) — I vividly remember, um, the story of another guest who was in aviation and basically the same story. Corona broke out, nobody picked up the phone anymore, so, Hmm. Yeah, you have not been alone, but it’s, it’s a tough time for all the startups hit there, um, really to survive. Um, and when you pivoted when you went into financial services, that’s the point. It makes sense for Ingo and St. Paul’s to come in. Ingo, can you tell us a little bit about first, what the corporation was all about with mines medical?
Ingo (00:05:17) — Yeah, I can, uh, Lucas mentioned it already, so it was a specific solution for insurers, um, to, uh, classify and extract, um, um, medical information from documents and as you were, as you were referring to Corona, a funny story on my side, and you, you have listeners in, uh, in, um, in New York. Um, so I spent half of the time, uh, of Corona in New York, actually. And what we learned then was, um, you know, from one day to the other, um, you know, people didn’t pick up the phones and they couldn’t as any physical documents anymore. And, uh, so digitizing them and reading and understanding them became actually business critical, critical from one day to the other. And that was true in banking, you know, for let’s say heavy compliance processes and for insurance as well. And I think this is really a global story, and I mentioned at the beginning we are a global company and we believe what, uh, Lucas and his team of mine’s medical, uh, you know, brought to the table as something that has, uh, global potential. And it is about, um, has said understanding what is in documents unstructured and helping, um, key roles, expert roles within insurance companies like underwriters, like claims adjusters to do their job, but even better, right? So concentrate on understanding what is, uh, in the document and not just reading through lengthy, uh, pages.
Joe (00:07:01) — We may take a few steps back because depending on the podcast, there is, um, sometimes 60 to 80% of our audits not from Germany. And we may explain that Germans love to be insured on everything. If you do have this kind of insurance on a car, it’s called FO because everything is paid for, and therefore it’s referred to in German as Foco mentally. So a lot of Germans do like to do a lot of insurance. That’s why it’s a, not only a big business here, but it’s a big business in Germany. And that’s why there’s, um, a, a need for you guys. Um, Ingo you refer to, um, underwriters that are basically the people in the insurance company. They take a look at the, at the case, at the document, at the person who wants insurance, and they either say, go or no go. Basically, that’s their job. And that’s where formally mine’s medical now, sun Pulse has eight, comes in getting data from unstructured data into really structured data. Um, Lucas, what kind of information are we talking about here to support the forecast co mentality data? ,
Lukas (00:08:21) — . So that’s, uh, unstructured data, especially in the, in the healthcare part comes in uncounted forms. Um, this, this can be semi-structured questionnaires where people, um, hand write some information at checking boxes, or it can be doctor’s letters, uh, or letters or all kind of medical discharge letters, all these kind of things. And to give you an example why, um, to extract these information is, is quite difficult and can’t be hard programmed into a software is that, um, one university hospital in Germany, um, I’ve been working with has hundred 70 different forms for a discharging person, people, hundred 70 forms, uh, word formats for, for that. So they come in a gazillion different, different types. So you need, you need software and you need machine learning, natural language processing to capture that information and actually understand what’s in there. Um, uh, and you, you can’t work there with some basic ocr, um, and, uh, yeah, forms basically. So it, it comes in all kinds of forms and we, we are talking, um, medical language, we are talking lay language as well, so people describing, especially in the insurance industry, like, um, if I’m describing my heart disease, uh, in a certain way, I’m certainly gonna do to describe it as a, uh, professor, Dr. Something who is just talking about that. Um, or I’m gonna say, oh, my back hurts. And somebody, uh, from the, uh, medical part would describe it early, fairly differently. So to, to capture that information,
Joe (00:10:07) — Then it is called lobar back pain. Hmm.
Lukas (00:10:11) — , right? . So, um, so with, with these kind of difficulties, we have to work on a daily basis with our software because these underwriters get this, get this information mostly in an email with, uh, just free text already, then questionnaires and all these attachments that they have collected over the years from different doctors from different, different parts. And, um, this problem is only gonna increase in the future when we have the electronic, uh, patient file where they’re not only collect, uh, the data they happen to keep in the last 10 years, or that they find somewhere and take a photo of it, but the collected health data in that form is gonna be collected in an online repository, , where then hundreds and hundreds of these unstructured data forms, uh, will be presented to the underwriter and they will be like, oh, now I have to read that. How do I do that? with a couple of people in my department.
Joe (00:11:07) — I see the business case. I, I, I’ll, I’ll soon get to INGO again. Um, but I, I’ve been doing some, uh, research while you’ve been speaking. There are more than 1,880 hospitals right now in Germany. And if they all have 17 discharge forms, uh, we are talking about more than 30,000 different form pages here just to discharge patients, just in order to give the people here and inside what size we are talking about here. Ingo is already smiling. Um, and the thing is, what you are referring to is Germany is moving slowly into a phase of keeping all medical data on a trip on your, uh, health insurance card. So Germany has a legal requirement to have health insurance that usually goes along with a cart with a chip on it. And with this, you can give your doctor access to all the data before, because before that it used to be the case, oh, you have to fill out like 17, 80, 90 20 pages, uh, because the doctors want to start to understand the patient. And that now changes with this, um, how is it, how is it officially called?
Lukas (00:12:25) — Electronic patient? Electronic .
Joe (00:12:29) — Ah, okay. And then it totally makes sense if you guys can make a structured data out of that. And I do believe that is where I hand over to Ingo who will tell us about the business case and how awesome this will be. Oh,
Ingo (00:12:46) — Yeah. Um, it is, and as I mentioned, we, we strongly believe in, in, um, let’s say the transformation and the accelerated digitization that is going on in the insurance space and the health space everywhere. And so, um, when we partnered with Lucas and his team, you know, we built a joint offering very specific for dis disability insurance in Germany, but we learned that this is actually something as said, generic, you know, you can, um, improve the situation for everybody, by the way, for, um, the sender, for the receiver, um, uh, because you are quicker in processing the documents, you, you, uh, you know, make better,
Joe (00:13:34) — Uh, what Ingo refers to is actually an occupational disability insurance, Lucas, because I, I’m not really sure if this is existing in any other country out there, maybe, uh, Austrians, Switzerland as well, but many people are listening.
Lukas (00:13:52) — It is, it is a very German, very German specific, uh, thing as, as you say, Paulka mentality. Um, but, uh, it’s, it’s a great, it’s a great insurance and about 30 per 30% of the people in Germany have that sort of insurance already, um, because it protects them from, uh, disability during their, uh, working lifetime and, um, yeah, helps, helps them feed their family should something unexpected happen like a disease and accident or whatever. Um, so they, they can do that and, uh, to sell these insurances, brokers are used, um, and they bring this information to the, to the insurance companies. So that’s that. But I think what Ingo was going at is that building this offering for insurance companies in a very specific offering German market, um, with Simple eight, um, we are building a product portfolio, uh, a product ecosystem for the insurance market, for a global insurance and financial industry market where the intelligent document processing is only one small part of, of that whole offering we bring to, to the market.
Lukas (00:15:03) — And we want to expand our portfolio, our technologies, our technologies to financial services and insurance companies all over the world. So it really is about scaling mine’s medical focused on a German market and, and scaling that to a global audience in insurance and, uh, and financial, financial services, and leaving that little little German space of health insurance and life insurance and bring this to, uh, banks, other insurance reinsurance on a, on a global scale. And, uh, integrating that into the Pulse eight ecosystem, product ecosystem where we’ll be really a full, full stop shop, um, consulting and technology company with Simples and Simples eight. So mm-hmm. , um, this is, this is bigger than Mine’s medical, obviously. Mm-hmm. ,
Joe (00:15:55) — And, and, and don’t get me wrong, I totally understand why you could not do this with, for people. Um, and of course, go going a little bit more general, totally make sense of, by the way, I just checked and in 2021, more than 14 million people in Germany had such an occupational disability insurance. Um, so that totally makes sense. So, um, ca c ca can you help us a little bit to understand how this acquisition process actually worked? Um, I have in mind, uh, you Lucas and Ingo sitting, uh, in the evening over beer, um, and he ask you, Hey, can I buy you a beer or your company
Lukas (00:16:35) — ? Pretty, pretty much. Um, uh, we, we met, first we met 2017, uh, at the Munich, Munich train, main train station. And, uh, that was the first meeting. And, uh, Ingle was kind enough to buy me a coffee, at least before, before asking that. No, I’m kidding. Um, and we have, we have been in contact and, uh, doing RFPs together for, uh, healthcare and, and things together. So this was in, in no means a cold call, Ingo, Hey, uh, I’d like to buy you coffee and your company and, uh, let’s see what, where this goes. It has been, uh, a year’s long partnership, um, with, with getting to know each other, how we operate, what kind of people we are, because mergers and acquisitions in this thing, in these, in this space, in my opinion, only work out in the, in the long term if you, if the founders, uh, want to be kept invested in it, if it works on a personal level.
Lukas (00:17:37) — And the post merger integration, uh, works that the, that the people in the company that the culture is a fit. And, uh, we were lucky enough that we could figure out that this culture is so similar and this is, uh, is set on a growth mindset, um, and on a personal level that we, that we can do that. And obviously then, um, yeah, , maybe you can also join in from your perspective, but from my perspective, it was a very natural thing when, when you approached us and say, Hey, could you, the partnership is go going great, but could you think about what, what do you think about if we extend that if you, if you make that more formalized partnership? And for me, that was a very natural, uh, question to come and, um, yeah, I was very happy that didn’t go, asked their questions. Um, the whole team, the whole team joined, uh, Simple as, now I’m, I’m not in liberty to say,
Joe (00:18:34) — Oh, too bad, too bad,
Lukas (00:18:35) — But, uh, but a hundred percent of mine’s medical got acquired.
Joe (00:18:39) — A hundred percent of mine’s medical got acquired. Okay, I see. Um, thing is for me right now, um, I’m interested in how you guys are planning to Concord World,
Lukas (00:18:53) — .
Joe (00:18:53) — What are the next steps you guys are, uh, you guys will be doing, um, here together with Sin Pauls? And by the way, your team, your team remains here and has mind, right?
Lukas (00:19:07) — That’s right. Um, simples is set up as a remote, or especially as a, is set up as a global and, uh, remote first company. So obviously we have still the Minds medical office in Frankfurt where the team that remains in Frankfurt regularly meets. But, uh, the main, the main office from Simples, Germany is in Disseldorf. Uh, but we are a global company with our headquarters in Switzerland, and now I think 1,300, uh, people working globally with the huge growing market in the apac APAC area, but also expanding to the United States. And as eight already has, I think 600 people working that, and we’re aiming, aiming to grow another, I don’t know, 30, 40, 50%, uh, this year alone. So mm-hmm. , um, and that, that’s actually part of the, the plan to conquer the world where I said earlier that, um, mine’s medical’s, nucleus of machine learning expertise and products is only a very small, small part of, of the offering that is, that is being brought to the, to the market because we really much very much believe that our client and our client base, once the stability Inga was referring to not only have, um, uh, consultants working there and producing slides and pro producing management advice, but also building products that scale that are products that are, um, hosted by somebody who reliable.
Lukas (00:20:41) — And this is basically what we are doing. I have been spending the last three months since the acquisition to expanding the team. We are now working with, uh, uh, great, great stable team within Simples eight, which is in Bratislava and in Frankfurt and in and the area, and working together and expanding our product vision, uh, building our first products, um, or stabilizing our first products, uh, also for the financial industry. So when we talk about intelligent document processing, we think about core architecture that is able to process any kind of documents, and then applying use case after use case after use case to that architecture. Um, which the first use case is already being implemented and, uh, working at, at customers, but then we see, okay, something new is coming up, something in a different market that is scalable, that is, uh, that brings huge benefit to, to the customer.
Lukas (00:21:41) — And then we create a new module that is plucked into that. And then we have a new use case, so we can easily scale the core of our application to new use cases. Obviously, always a little bit of retuning the models, uh, the machine learning models retuning that, retraining that a little bit. Um, for, for a new use case, because it’s every time a little bit different, but at, at its core, we always have this, uh, document processing pipeline where some form of unstructured document comes in, is processed by humans, um, and put into an, uh, operating model, uh, for an automated process. And, uh, the, the key gap there is the humans being overwhelmed with thousands and thousands of documents in many different different verticals. And, um, for our specific clients in insurance and financial services, we were gonna help, uh, be that little part in the Pulse eight strategy to help them, uh, have better processes and, um, enable them to do what they’re really good at. Mm-hmm. and, and not reading a piece of paper that has been read five times already.
Joe (00:22:54) — I see. Uh, what comes to mind, uh, from my personal experience is, for example, when you want to open an account, um, for a legal entity, so you have to get all the paperwork for the legal entity, all the paperwork for the people who are actually working on behalf of this entity, maybe the owners and so on and so forth. So that, quite a lot of stuff you have to do. That was one of the use cases that came to mind for me. Um, Ingel, what kind of other use cases do you have in mind? Unfortunately, we lost him in the parking lot again, , we really, sorry, guys here. Um, that is, that is a little bit, um, uncommon here, but, uh, Lucas stays with us, uh, by the way, um, big mental high five that he is here because he’s now, um, in a house, uh, owned by his parents in law, uh, because he and his family do have Corona and therefore, um, they, uh, he is still doing this interview. Awesome. Thank you very much. Great.
Lukas (00:24:00) — Of course for you. Anytime. Anytime. Joe .
Joe (00:24:03) — Great. Um, you, you guys have been headquartered here in Frankfurt, Rhine, mine, you got incorporated in 2019. Did you exit in very early 2023? Uh, what, what kind of support, what was useful for you in this area, uh, through your entrepreneurial journey from the, from the startup, from start of the startup until the end
Lukas (00:24:27) — Incorporation was 2016 already 16.
Joe (00:24:30) — Oh, I’m sorry.
Lukas (00:24:31) — No, no worries. No worries. Um, uh, instrumental for, for our success, um, were many, many projects that helped us get off the ground, the Exist project. Um, many, many German listeners or in the entrepreneurial space will know that it is a, um, startup grant basically grant granted to entrepreneurs, uh, getting, getting off the ground. We were able to, to secure this grant in the beginning mm-hmm. exist, um, to, to yeah, quit our jobs, our day jobs and focus a hundred percent on the company. Without that, we probably wouldn’t have been able to do that, um, for the first year. Um, then o obviously the, we, we were awarded with the Frank futa, founder, founder, uh, prize, first prize in 2018 for what we did. Uh, also the, um, innovation, the first, the innovation prize for, uh, the, he, he prize has a founder award in 2018 as well.
Lukas (00:25:37) — So 2018 was a great year for us, um, which also helped, uh, to get funding. Uh, our first funding was 2018, so 2002 years after incorporation mm-hmm. . Um, we, we got the, uh, first funding with first, uh, small round with Business Angels. And then, uh, also in 2018, later in the later part of 2018, with the, uh, bha, uh, Heile three, uh, founding, or what’s it called, um, project, uh, where the state of Hessan invests in promising startups together with bu with Business Angels. And so you could say the state of Hassan, uh, was very much, uh, instrumental for our success so we could, uh, grow and build our products for the first years. And, um, and do that, you know, this, this was, uh, this was very helpful in the first years. Mm-hmm.
Joe (00:26:34) — , I see.
Lukas (00:26:35) — I’m also also a, uh, r and d grant that was granted the Luva project, Luva three project mm-hmm. , where we were able to, um, do r and d together with the universities, um, and, uh, do our research and project or product development in, in the space of transfer learning.
Joe (00:26:54) — Great. Um, it’s always important to have, um, entrepreneurs, experience entrepreneurs as mentors, uh, help you stay there, here in the local startup scene. And the very last question, if you would ha, would be asked about one or two suggestions, how to improve the local startup system, what would it be?
Lukas (00:27:16) — So, I, I think there are great, great activities already going on with the uni barto from the, from the GTU University or the, um, all the other activities that, that are going on. And that really help, um, sometimes I feel that, especially for young founders or first time founders who have really no idea where, where to start some, something where they could better find these informations, uh, or have, have some mentor at hand, um, that, that’s always very helpful. But at the end of the day, guys, it’s your startup. You want to get it off the ground. So, um, do the legwork, ask the people and do that. It’s, it’s not something somebody’s just offering you. If you, if you are not able to build your own network and, and to find these things , then that’s on you, on you founders. So, um, that’s the easy part of found of founding a company. The hard, hard work comes after that. So I think the startup ecosystems set up pretty well and the help is there if you want it, but go get it. Uh, there are great people, great mentors that, that are willing to help you and put, put the, put the money where their mouth is and, um, but you have to shine and do the legwork you have to
Joe (00:28:34) — Try. Awesome. Great, uh, amazing closing words. Thank you Lucas, for being here. And maybe Engel can also say his goodbye. He joined us. I, I do believe for the fourth or fifth times, , he always gets kicked out. Thank you for also, uh, being here, um, our guest as well and doing as much ex explaining as you can do.
Lukas (00:28:57) — Thank you, Joe. And, uh, maybe up to a third time we’ll see . No, I’m kidding. I’m very happy with Simple, but you never know. .
Joe (00:29:05) — Maybe as an investor.
Lukas (00:29:07) — Maybe as an investor, we’ll see . All right,
Narrator (00:29:16) — That’s all folks. Find more news streams, events and interviews at www io. Remember, Sherry is Carrie.
