1 00:00:00,080 --> 00:00:02,179 Hello and welcome to the Becker's Healthcare podcast. 2 00:00:02,240 --> 00:00:04,000 My name is Chanel Banger and I'm recording 3 00:00:04,000 --> 00:00:06,099 live at the Becker's Healthcare future of dentistry 4 00:00:06,160 --> 00:00:06,660 roundtable, 5 00:00:07,040 --> 00:00:09,759 sitting down with Rob Giordina and doctor Ellen 6 00:00:09,759 --> 00:00:12,820 Polsky. And today we're gonna discuss physio 360 7 00:00:12,960 --> 00:00:14,099 and AI in dentistry. 8 00:00:14,400 --> 00:00:15,974 But before we get into that, could I 9 00:00:15,974 --> 00:00:18,614 please have you both introduce yourselves? Doctor Polsky, 10 00:00:18,614 --> 00:00:19,594 I'll start with you. 11 00:00:20,135 --> 00:00:23,335 Yes. Hi, Ellen Polsky. I am a CEO 12 00:00:23,335 --> 00:00:26,375 of, one of the DSOs. We have about 13 00:00:26,375 --> 00:00:27,195 30 locations, 14 00:00:27,654 --> 00:00:28,794 across the country. 15 00:00:29,414 --> 00:00:32,260 And I'm by training a pediatric dentist, 16 00:00:33,280 --> 00:00:35,380 and this is basically the background. 17 00:00:36,159 --> 00:00:38,880 Thank you. And Rob? Hi. I'm Rob Giordino. 18 00:00:38,880 --> 00:00:40,899 Thanks for having us. I'm a technologist, 19 00:00:41,439 --> 00:00:42,880 about 30 years of, 20 00:00:43,359 --> 00:00:45,914 experience in in the data and software industry. 21 00:00:46,875 --> 00:00:49,215 Couple big companies like like Amazon and Oracle, 22 00:00:49,354 --> 00:00:50,954 most recently, Palantir, and now, 23 00:00:51,835 --> 00:00:54,795 4 years working on our new technology, which 24 00:00:54,795 --> 00:00:56,335 I'm excited to talk about today. 25 00:00:56,635 --> 00:00:59,100 Well, definitely. Let's get right into it then. 26 00:00:59,340 --> 00:01:00,780 Could you tell us a bit more about 27 00:01:00,780 --> 00:01:02,559 how this AI relates to dentistry? 28 00:01:03,259 --> 00:01:06,140 Right. So it's interesting to talk about which 29 00:01:06,140 --> 00:01:08,479 kind of AI we're gonna talk about first. 30 00:01:08,859 --> 00:01:11,019 So what we're talking about today is not 31 00:01:11,019 --> 00:01:13,420 so much the traditional health care AI for 32 00:01:13,420 --> 00:01:15,795 focusing on diagnosis. So we're not talking about 33 00:01:15,795 --> 00:01:17,875 examining image imagery or trying to come to 34 00:01:17,875 --> 00:01:19,015 a complicated diagnosis. 35 00:01:19,555 --> 00:01:22,515 That's an interesting problem and really valuable, but 36 00:01:22,515 --> 00:01:23,954 we all know that that's a work in 37 00:01:23,954 --> 00:01:24,454 progress. 38 00:01:24,914 --> 00:01:26,594 So, so we're, we're actually talking about the 39 00:01:26,594 --> 00:01:28,380 other AI. So what we wanna talk 40 00:01:28,760 --> 00:01:31,240 about today is the revolution inspired by Chat 41 00:01:31,240 --> 00:01:33,719 GPT and how that relates to the dental 42 00:01:33,719 --> 00:01:34,219 world. 43 00:01:35,159 --> 00:01:37,320 Amazing. I'm excited to dig into this with 44 00:01:37,320 --> 00:01:39,159 you both. Now, can you tell me the 45 00:01:39,159 --> 00:01:41,719 challenges associated with bringing this technology into the 46 00:01:41,719 --> 00:01:42,219 industry? 47 00:01:43,015 --> 00:01:44,954 Well, usually, what we experience 48 00:01:45,655 --> 00:01:47,734 in the dental field or any health care 49 00:01:47,734 --> 00:01:48,234 field, 50 00:01:48,694 --> 00:01:52,295 document to document solutions already exist, meaning we 51 00:01:52,295 --> 00:01:55,275 can put some kind of document into ChargeGDP, 52 00:01:55,495 --> 00:01:57,355 and ChargeGDP can analyze it. 53 00:01:57,980 --> 00:02:01,020 However, in health care, we do use, 54 00:02:01,340 --> 00:02:03,520 very different types of solution, business 55 00:02:04,140 --> 00:02:04,640 applications, 56 00:02:06,140 --> 00:02:06,640 operational 57 00:02:06,939 --> 00:02:08,000 type of applications. 58 00:02:08,460 --> 00:02:10,560 So the data is very complicated 59 00:02:10,860 --> 00:02:13,775 and not in a document form. 60 00:02:14,314 --> 00:02:16,414 So that presents significant challenge 61 00:02:16,794 --> 00:02:19,375 in terms of data sorting and data organization. 62 00:02:20,875 --> 00:02:22,875 Got it. And when you talk about this 63 00:02:22,875 --> 00:02:26,189 data, do you mean like EHR data, prescriptions? 64 00:02:27,050 --> 00:02:28,909 What all does this incorporate exactly? 65 00:02:29,610 --> 00:02:31,530 Yeah. So the vision there is you really 66 00:02:31,530 --> 00:02:34,489 wanna ask any kind of question. No matter 67 00:02:34,489 --> 00:02:36,729 what system part of the answer might live 68 00:02:36,729 --> 00:02:38,830 in, you wanna ask a question that covers 69 00:02:39,155 --> 00:02:39,814 things like, 70 00:02:40,194 --> 00:02:42,435 electronic health records or medical records or practice 71 00:02:42,435 --> 00:02:44,835 management system. And it might involve some of 72 00:02:44,835 --> 00:02:46,855 the staff. So you have an HR system. 73 00:02:47,235 --> 00:02:49,715 It might involve calls or accounting, and you 74 00:02:49,715 --> 00:02:52,080 really don't want to see those boundaries when 75 00:02:52,080 --> 00:02:53,599 you're asking the question, and you want an 76 00:02:53,599 --> 00:02:55,540 answer that can pull data from all of 77 00:02:56,080 --> 00:02:56,819 them. Absolutely. 78 00:02:57,200 --> 00:02:58,879 And can you walk us through how this 79 00:02:58,879 --> 00:03:01,680 AI does the work for you? Okay. So 80 00:03:01,680 --> 00:03:04,275 when you have all these disconnected systems, you're 81 00:03:04,275 --> 00:03:06,034 really asking a little bit of your questions 82 00:03:06,034 --> 00:03:07,474 out of each one. So you have to 83 00:03:07,474 --> 00:03:08,995 you have to check with your accounting system. 84 00:03:08,995 --> 00:03:10,675 You have to go into your patient management 85 00:03:10,675 --> 00:03:13,314 system. So the idea here is that if 86 00:03:13,314 --> 00:03:15,635 you pull them all together in one model 87 00:03:15,635 --> 00:03:17,650 that makes sense to the business, things that 88 00:03:17,650 --> 00:03:19,330 are obviously the way you would draw them 89 00:03:19,330 --> 00:03:20,710 on a whiteboard, for example. 90 00:03:21,250 --> 00:03:23,569 This is easy for you to combine all 91 00:03:23,569 --> 00:03:25,889 your systems into that model, particularly with today's 92 00:03:25,889 --> 00:03:28,150 AI. But it is also extremely 93 00:03:28,610 --> 00:03:29,110 understandable. 94 00:03:29,569 --> 00:03:31,409 That kind of whiteboard model is the kind 95 00:03:31,409 --> 00:03:33,075 of thing that an AI, a chat gbt 96 00:03:33,075 --> 00:03:33,735 style AI 97 00:03:34,275 --> 00:03:36,455 can answer natural questions about. 98 00:03:37,395 --> 00:03:39,474 Love to hear it. And doctor Polsky, could 99 00:03:39,474 --> 00:03:41,955 you talk about exactly what you're doing today 100 00:03:41,955 --> 00:03:42,935 with this technology? 101 00:03:43,795 --> 00:03:46,550 Today, we're trying to set an interconnected system, 102 00:03:46,610 --> 00:03:48,530 and one of the aspects that we used 103 00:03:48,530 --> 00:03:50,069 to struggle with that manual 104 00:03:50,689 --> 00:03:53,110 is absolutely terrible and horrible. 105 00:03:53,490 --> 00:03:56,050 Any type of reports that any of my 106 00:03:56,050 --> 00:03:58,230 staff members were trying to pull out manually 107 00:03:58,450 --> 00:04:00,205 was requiring a lot of labor. 108 00:04:00,584 --> 00:04:03,324 And that part, I think this particular technology 109 00:04:03,384 --> 00:04:05,965 can solve very beautifully and elegantly for us. 110 00:04:06,185 --> 00:04:09,145 So in many situations when the company tries 111 00:04:09,145 --> 00:04:10,764 to pull out the data manually, 112 00:04:11,305 --> 00:04:14,199 you end up in a scenario where you're 113 00:04:14,259 --> 00:04:15,879 acting very reactively. 114 00:04:16,339 --> 00:04:18,979 Somebody urgently needs the data. They pull all 115 00:04:18,979 --> 00:04:21,459 the resources. They try to put 5 people 116 00:04:21,459 --> 00:04:23,379 or 10 people on a task and pull 117 00:04:23,379 --> 00:04:25,379 that data manually, and then they kind of 118 00:04:25,379 --> 00:04:28,204 go and forget it. So what this solution 119 00:04:28,204 --> 00:04:29,104 does, it 120 00:04:29,404 --> 00:04:29,904 crisply 121 00:04:30,284 --> 00:04:33,024 defines that particular problem as 122 00:04:33,404 --> 00:04:35,404 I can ask any questions for my data, 123 00:04:35,404 --> 00:04:37,904 and now no manual resources are necessary 124 00:04:38,284 --> 00:04:41,664 to answer the questions all across the system. 125 00:04:42,120 --> 00:04:43,560 Got it. And that brings me to my 126 00:04:43,560 --> 00:04:45,639 next question. How do you put this all 127 00:04:45,639 --> 00:04:47,720 together, and what is your new vision with 128 00:04:47,720 --> 00:04:48,220 this? 129 00:04:48,759 --> 00:04:51,560 Okay. Sure. So that's a big question because 130 00:04:51,560 --> 00:04:53,800 we're we're changing that future. We're evolving it 131 00:04:53,800 --> 00:04:55,480 now. But I think we can go back 132 00:04:55,480 --> 00:04:57,819 to what everybody really wants about their practice. 133 00:04:58,335 --> 00:04:59,074 They really 134 00:04:59,455 --> 00:05:01,055 like, the best decision that you can make 135 00:05:01,055 --> 00:05:02,574 about a difficult issue is when you have 136 00:05:02,574 --> 00:05:04,735 the perfect spreadsheet in front of you. All 137 00:05:04,735 --> 00:05:06,595 the exact right numbers, nothing 138 00:05:06,895 --> 00:05:08,975 irrelevant, just so that you can understand a 139 00:05:08,975 --> 00:05:10,870 pattern or make a make the best possible 140 00:05:10,870 --> 00:05:12,870 decision. So that's where we're going. We wanna 141 00:05:12,870 --> 00:05:14,550 always be able to have and assemble just 142 00:05:14,550 --> 00:05:16,069 the right amount of data for us for 143 00:05:16,069 --> 00:05:18,470 for today, which requires that, as we talked 144 00:05:18,470 --> 00:05:19,829 about earlier, you have to be able to 145 00:05:19,829 --> 00:05:22,329 interconnect all your data. It can't be unavailable 146 00:05:22,389 --> 00:05:23,910 just because a little bit of the data 147 00:05:23,910 --> 00:05:25,615 is in a different system. So you also 148 00:05:25,615 --> 00:05:27,375 want that to be automated so that it's 149 00:05:27,375 --> 00:05:29,455 always working for you on a clock and 150 00:05:29,455 --> 00:05:30,435 always up to date. 151 00:05:30,895 --> 00:05:33,055 And clearly, as Ellen said, you just wanna 152 00:05:33,055 --> 00:05:34,975 be able to answer any question, any reasonable 153 00:05:34,975 --> 00:05:36,735 question. You know what's in the data. You 154 00:05:36,735 --> 00:05:38,675 just wanna be able to answer it fluidly 155 00:05:38,735 --> 00:05:40,899 as your needs change day by day. Thank 156 00:05:40,899 --> 00:05:42,680 you so much for walking us through that. 157 00:05:42,819 --> 00:05:44,579 And I'll turn it back to you, doctor 158 00:05:44,579 --> 00:05:46,819 Polsky. What value would you say this brings 159 00:05:46,819 --> 00:05:47,560 to the customer? 160 00:05:48,019 --> 00:05:50,279 Well, the biggest value it brings, my CEO 161 00:05:50,500 --> 00:05:51,240 used to 162 00:05:51,699 --> 00:05:54,500 spend about 2 hours daily trying to pull 163 00:05:54,500 --> 00:05:57,644 the reports together from various data sources to 164 00:05:57,644 --> 00:05:59,345 combine into 1 Excel spreadsheet 165 00:06:00,125 --> 00:06:03,664 and also occupying valuable resources for that particular 166 00:06:03,724 --> 00:06:06,384 task. So reduction in labor cost is absolutely 167 00:06:06,524 --> 00:06:09,120 one of the biggest valuable parts that we 168 00:06:09,339 --> 00:06:10,160 see. Secondly, 169 00:06:10,620 --> 00:06:12,399 since the data is so interconnected, 170 00:06:12,699 --> 00:06:15,100 we actually can trace the patient outcomes, which 171 00:06:15,100 --> 00:06:16,860 in medicine, we are starting to see in 172 00:06:16,860 --> 00:06:19,519 the solutions. In terms of dental, we don't. 173 00:06:19,819 --> 00:06:20,319 So 174 00:06:20,745 --> 00:06:21,564 patient outcomes 175 00:06:22,345 --> 00:06:25,145 are improving because of this type of data 176 00:06:25,145 --> 00:06:27,004 pools. And thirdly, 177 00:06:27,384 --> 00:06:29,404 we are able to react proactively 178 00:06:30,024 --> 00:06:33,165 towards the problems, not being in constant reactive 179 00:06:33,225 --> 00:06:36,240 mode, which is less stress on my employees, 180 00:06:36,699 --> 00:06:38,319 much more relaxed environment, 181 00:06:38,860 --> 00:06:41,920 and now we're planning ahead versus constantly reacting. 182 00:06:42,139 --> 00:06:44,300 So we kind of all know the best 183 00:06:44,300 --> 00:06:46,319 algorithms for running our practices, 184 00:06:46,779 --> 00:06:49,415 but we just don't have that cohesive data 185 00:06:49,415 --> 00:06:51,175 layer to deploy it. So I think let's 186 00:06:51,175 --> 00:06:53,355 kind of get together and start doing it. 187 00:06:53,894 --> 00:06:56,134 Got it. I love the interoperability of all 188 00:06:56,134 --> 00:06:58,694 this. Well, Rob and doctor Polsky, I wanna 189 00:06:58,694 --> 00:07:00,535 thank you for your time today. But before 190 00:07:00,535 --> 00:07:02,134 I let you go, is there anything else 191 00:07:02,134 --> 00:07:03,355 that listeners should know? 192 00:07:03,790 --> 00:07:05,470 I think Ellen said it really well, and 193 00:07:05,470 --> 00:07:08,509 that's the final word, is that everybody knows 194 00:07:08,509 --> 00:07:11,069 the algorithms they wanna use to optimize their 195 00:07:11,069 --> 00:07:13,389 business. The data stands in the way, and 196 00:07:13,389 --> 00:07:14,750 it shouldn't. You should get to a place 197 00:07:14,750 --> 00:07:16,189 where you have it fully connected and you 198 00:07:16,189 --> 00:07:17,970 can put your best ideas into practice. 199 00:07:18,574 --> 00:07:20,814 Love to hear it. Well, doctor Polsky and 200 00:07:20,814 --> 00:07:22,415 Rob, I wanna thank you once again for 201 00:07:22,415 --> 00:07:24,014 your time today and for sharing your insights 202 00:07:24,014 --> 00:07:26,334 on the Becker's Healthcare podcast. Thank you. It's 203 00:07:26,334 --> 00:07:28,035 great to be here. Thank you.