1 00:00:02,560 --> 00:00:05,299 Today's clinical providers and healthcare systems 2 00:00:05,599 --> 00:00:08,400 need a strategic partner that is focused on 3 00:00:08,400 --> 00:00:10,339 the evolution of healthcare delivery. 4 00:00:10,960 --> 00:00:14,080 At Surgery Partners, we are redefining the healthcare 5 00:00:14,080 --> 00:00:14,414 industry 6 00:00:14,894 --> 00:00:18,195 as a nation's leading independent operator of surgical 7 00:00:18,335 --> 00:00:20,434 facilities and ancillary services. 8 00:00:20,894 --> 00:00:24,574 With an extensive presence spanning over 180 9 00:00:24,574 --> 00:00:25,714 locations nationwide, 10 00:00:26,335 --> 00:00:28,515 our commitment extends beyond healthcare. 11 00:00:29,079 --> 00:00:31,339 It's about fostering successful partnerships 12 00:00:31,800 --> 00:00:33,799 that enhance the quality of care in the 13 00:00:33,799 --> 00:00:35,100 communities we serve. 14 00:00:35,640 --> 00:00:38,280 Surgery Partners is more than an operator or 15 00:00:38,280 --> 00:00:41,880 service provider. We are your strategic ally committed 16 00:00:41,880 --> 00:00:43,844 to transforming healthcare delivery. 17 00:00:44,565 --> 00:00:47,545 Learn more at surgerypartners.com. 18 00:00:50,005 --> 00:00:52,245 This is Laura Dierda with the Becker's Healthcare 19 00:00:52,245 --> 00:00:54,325 podcast. I'm thrilled today to be joined by 20 00:00:54,325 --> 00:00:55,625 doctor Ashish Atresha, 21 00:00:56,020 --> 00:00:58,039 who's a professor of medicine and gastroenterology 22 00:00:58,420 --> 00:01:00,100 at UC Davis Health, as well as a 23 00:01:00,100 --> 00:01:03,060 venture partner with Global Ventures X. Doctor Atresha, 24 00:01:03,060 --> 00:01:04,020 it's a pleasure to have you on the 25 00:01:04,020 --> 00:01:05,000 podcast today. 26 00:01:05,299 --> 00:01:07,239 Thank you so much, Laura, for having me. 27 00:01:07,379 --> 00:01:09,540 Now I I'm excited for our conversation because 28 00:01:09,540 --> 00:01:11,284 I know you've had quite a a great 29 00:01:11,284 --> 00:01:13,784 career and a lot of different experiences. So 30 00:01:14,084 --> 00:01:16,084 I'm wondering, first and foremost, can you introduce 31 00:01:16,084 --> 00:01:17,605 yourself and tell us a little bit about 32 00:01:17,605 --> 00:01:18,344 your background? 33 00:01:19,045 --> 00:01:22,245 Yeah. Happy to. In fact, I, often say 34 00:01:22,245 --> 00:01:24,084 that this is my fifth career life. I'm 35 00:01:24,084 --> 00:01:26,185 getting into very lucky to have that. 36 00:01:26,730 --> 00:01:28,969 And my career journey started as a physician 37 00:01:28,969 --> 00:01:30,730 and a public health scholar. I did master's 38 00:01:30,730 --> 00:01:33,290 in public health and then residency in Cleveland 39 00:01:33,290 --> 00:01:34,269 Clinic. And, 40 00:01:34,569 --> 00:01:37,049 because of my experiences both on the clinical 41 00:01:37,049 --> 00:01:37,549 side, 42 00:01:38,344 --> 00:01:40,185 as well as on the research side, we 43 00:01:40,185 --> 00:01:41,864 were the first ones in Cleveland Clinic to 44 00:01:41,864 --> 00:01:42,364 implement 45 00:01:42,745 --> 00:01:44,204 Epic after Pfizer, 46 00:01:44,584 --> 00:01:46,344 and I could do my research study right 47 00:01:46,344 --> 00:01:47,864 there from the data that came out of 48 00:01:47,864 --> 00:01:51,165 it. And, so that prompted me actually to 49 00:01:51,290 --> 00:01:53,049 take the second part of my career, which 50 00:01:53,049 --> 00:01:53,790 was informatics. 51 00:01:54,729 --> 00:01:57,290 I did informatics fellowship and helped implement the 52 00:01:57,290 --> 00:01:57,790 inpatient, 53 00:01:58,489 --> 00:02:01,689 epic, at Cleveland Clinic and also associate program 54 00:02:01,689 --> 00:02:04,650 director for informatics. So so my journey went 55 00:02:04,650 --> 00:02:05,150 from 56 00:02:05,545 --> 00:02:08,344 seeing clinicians to being informatics to scale what 57 00:02:08,344 --> 00:02:11,245 we do and bring efficiency into that manner. 58 00:02:12,104 --> 00:02:14,665 And I quickly realized that, in my GI 59 00:02:14,665 --> 00:02:16,205 world as I was doing my fellowship, 60 00:02:17,224 --> 00:02:19,805 informatics was mostly at that stage covering 61 00:02:20,610 --> 00:02:23,509 electronic medical records, the physician part, the clinician 62 00:02:23,569 --> 00:02:24,069 part. 63 00:02:24,689 --> 00:02:26,129 But 99.9% 64 00:02:26,129 --> 00:02:27,889 of the time which patients spend is in 65 00:02:27,889 --> 00:02:29,030 their home or workplace, 66 00:02:29,569 --> 00:02:31,250 and we were not touching them. So all 67 00:02:31,250 --> 00:02:33,349 this chronic disease, all these health behaviors, 68 00:02:34,375 --> 00:02:36,534 everything that could have really made an impact 69 00:02:36,534 --> 00:02:37,675 on patient outcomes, 70 00:02:38,134 --> 00:02:40,055 we were not there as physicians guiding the 71 00:02:40,055 --> 00:02:40,555 patients. 72 00:02:41,175 --> 00:02:41,834 I remember, 73 00:02:42,294 --> 00:02:44,955 when I was managing patients with IBD, inflammatory 74 00:02:45,014 --> 00:02:46,875 bowel disease, they came from West Virginia, 75 00:02:47,310 --> 00:02:49,489 And this patient came with 20 bowel movements, 76 00:02:49,870 --> 00:02:50,530 for twenty 77 00:02:50,830 --> 00:02:53,150 years. And she was not managed properly, but 78 00:02:53,150 --> 00:02:55,229 she didn't have access to resources. And we 79 00:02:55,229 --> 00:02:56,909 didn't have a monitoring tool to know if 80 00:02:56,909 --> 00:02:57,889 we make a difference, 81 00:02:58,270 --> 00:02:59,810 did she get better or not. 82 00:03:00,185 --> 00:03:02,425 So I managed her. But when she went 83 00:03:02,425 --> 00:03:04,525 back, I realized six months later, 84 00:03:05,064 --> 00:03:06,824 she had a surgery, and there was a 85 00:03:06,824 --> 00:03:09,384 gap I felt in monitoring the patients at 86 00:03:09,384 --> 00:03:10,125 their home. 87 00:03:10,584 --> 00:03:11,564 So that started 88 00:03:12,264 --> 00:03:14,264 a big innovation drive in me, and I 89 00:03:14,264 --> 00:03:15,245 was lucky, meaning, 90 00:03:15,900 --> 00:03:16,560 to be, 91 00:03:17,980 --> 00:03:20,400 getting a word for innovation for virtual paging. 92 00:03:20,780 --> 00:03:23,099 So I knew that we, as clinicians, when 93 00:03:23,099 --> 00:03:24,159 we face a problem, 94 00:03:25,099 --> 00:03:27,120 we can, if we have the right tools, 95 00:03:27,500 --> 00:03:30,155 informatics and then digital health, we can be 96 00:03:30,155 --> 00:03:30,655 cocreators 97 00:03:30,955 --> 00:03:32,875 in problem solving. So a big part of 98 00:03:32,875 --> 00:03:34,735 my journey was going beyond informatics 99 00:03:35,354 --> 00:03:37,594 to really take on digital health. And I 100 00:03:37,594 --> 00:03:39,915 got recruited at Mount Sinai to be chief 101 00:03:39,915 --> 00:03:40,974 innovation officer 102 00:03:41,354 --> 00:03:42,014 in medicine 103 00:03:42,474 --> 00:03:45,000 where we established the App Lab, and that's 104 00:03:45,000 --> 00:03:47,379 where I've been called as an app doctor. 105 00:03:47,840 --> 00:03:50,099 So so we build apps across all specialty, 106 00:03:50,239 --> 00:03:52,739 basically solving problem of patient 107 00:03:53,039 --> 00:03:54,659 engagement and patient outcomes, 108 00:03:55,199 --> 00:03:57,759 the, what I call as the 99.9% 109 00:03:57,759 --> 00:03:58,500 of the problem. 110 00:03:59,014 --> 00:04:01,334 This had been amazing journey. We built the 111 00:04:01,334 --> 00:04:04,955 first platform to prescribe apps and AI bots, 112 00:04:05,414 --> 00:04:08,055 right from electronic medical record. We were able 113 00:04:08,055 --> 00:04:10,294 to pattern that, and then Mount Sinai was 114 00:04:10,294 --> 00:04:13,014 able to commercialize that. It ended up touching 115 00:04:13,014 --> 00:04:16,399 thirty to thirty million lives through health systems, 116 00:04:16,459 --> 00:04:19,419 and then UnitedCare was using for 60,000,000 patients. 117 00:04:19,419 --> 00:04:19,919 So 118 00:04:20,379 --> 00:04:22,479 that was amazing. As a doctor, 119 00:04:23,180 --> 00:04:25,100 one to one, if I would have provided 120 00:04:25,100 --> 00:04:27,339 care, I would have touched 5,000, 10 thousand 121 00:04:27,339 --> 00:04:28,479 patients in my life. 122 00:04:28,834 --> 00:04:30,214 But here through technology, 123 00:04:31,235 --> 00:04:33,235 we were able to touch, you know, 30 124 00:04:33,235 --> 00:04:34,375 to 90,000,000 patients. 125 00:04:34,834 --> 00:04:37,654 So UC Davis Health then hired me 126 00:04:38,274 --> 00:04:40,595 as a CIO and chief digital health officer, 127 00:04:40,595 --> 00:04:43,095 the inaugural chief digital health officer for UC 128 00:04:43,310 --> 00:04:43,810 System, 129 00:04:44,110 --> 00:04:46,430 basically to see how we bring innovation at 130 00:04:46,430 --> 00:04:48,210 scale really lead to transformation. 131 00:04:48,589 --> 00:04:50,430 And it has been amazing five four years 132 00:04:50,430 --> 00:04:50,930 there, 133 00:04:51,389 --> 00:04:54,029 where we made UC Davis the first health 134 00:04:54,029 --> 00:04:56,029 system in California to be digital health most 135 00:04:56,029 --> 00:04:59,444 wired level 10, partnered with AWS from cloud 136 00:04:59,444 --> 00:05:01,625 innovation center, a lot of cutting edge technology, 137 00:05:01,925 --> 00:05:04,585 including AI. But we started finding 138 00:05:05,205 --> 00:05:07,685 that generative AI coming two and a half 139 00:05:07,685 --> 00:05:10,405 years ago is just a completely different set 140 00:05:10,405 --> 00:05:12,105 of technology that we need to master, 141 00:05:12,639 --> 00:05:14,500 and no health system has 142 00:05:14,879 --> 00:05:16,800 the skills, talent, or resources to do it 143 00:05:16,800 --> 00:05:19,519 alone. So we formed a community called Valid 144 00:05:19,519 --> 00:05:21,300 AI, the South fifty eight organizations. 145 00:05:21,839 --> 00:05:23,279 And at the next part of my journey, 146 00:05:23,279 --> 00:05:26,319 I'm going beyond one organization to support multiple 147 00:05:26,319 --> 00:05:26,819 organization 148 00:05:27,245 --> 00:05:29,725 kind of, and as a venture partner to 149 00:05:29,725 --> 00:05:31,324 to kind of say, how do we really 150 00:05:31,324 --> 00:05:33,884 create a solve an ecosystem problem of health 151 00:05:33,884 --> 00:05:35,345 system, health plan, and pharma 152 00:05:35,725 --> 00:05:37,824 using generative AI as, 153 00:05:38,444 --> 00:05:40,064 as a tool and a platform, 154 00:05:40,680 --> 00:05:41,959 that can help us all if we do 155 00:05:41,959 --> 00:05:43,819 it the right way and bring convergence. 156 00:05:45,399 --> 00:05:47,319 That's amazing to hear. You know? And and 157 00:05:47,319 --> 00:05:49,240 really cool to hear, especially what you've been 158 00:05:49,240 --> 00:05:51,240 able to do, whether it's at Cleveland Clinic 159 00:05:51,240 --> 00:05:51,740 or, 160 00:05:52,199 --> 00:05:53,720 UC Davis Health and and, 161 00:05:54,354 --> 00:05:56,514 bringing generative AI. I I know a lot 162 00:05:56,514 --> 00:05:58,194 of hospitals and health systems right now are 163 00:05:58,194 --> 00:06:00,115 trying to troubleshoot that and figure out the 164 00:06:00,115 --> 00:06:02,694 best way that they can, incorporate the technology, 165 00:06:03,394 --> 00:06:05,414 given some of the challenges as well, 166 00:06:05,875 --> 00:06:07,889 in in a safe and secure manner. I'm 167 00:06:07,889 --> 00:06:10,209 curious from your perspective, what are the opportunities 168 00:06:10,209 --> 00:06:11,490 and headwinds that you have your eye on 169 00:06:11,490 --> 00:06:14,050 right now, especially knowing how many health systems 170 00:06:14,050 --> 00:06:15,729 are trying to figure out how to use 171 00:06:15,729 --> 00:06:16,709 AI appropriately? 172 00:06:17,729 --> 00:06:20,415 Yeah. Happy to answer. And we started our 173 00:06:20,495 --> 00:06:22,095 AI journey four, five years ago in a 174 00:06:22,095 --> 00:06:24,355 much more meaningful manner at UC Davis Health. 175 00:06:24,495 --> 00:06:26,574 And one of the things we are finding 176 00:06:26,574 --> 00:06:29,154 is, Laura, is this problem of plenty. 177 00:06:29,855 --> 00:06:31,395 But there's so many 178 00:06:32,975 --> 00:06:35,074 AI providers, like AI algorithm providers. 179 00:06:35,569 --> 00:06:38,050 There's so many cloud providers. There's so many 180 00:06:38,050 --> 00:06:38,550 products. 181 00:06:39,729 --> 00:06:42,069 Every or many of them have AI component, 182 00:06:42,289 --> 00:06:43,810 and there's so many solutions you can build 183 00:06:43,810 --> 00:06:45,829 on your own as well. So 184 00:06:46,930 --> 00:06:48,610 how do we choose what is the right 185 00:06:48,610 --> 00:06:50,884 one? And there is a fatigue happening 186 00:06:51,264 --> 00:06:53,345 because our resources are getting more and more 187 00:06:53,345 --> 00:06:55,904 limited, especially with the federal indirect cuts that's 188 00:06:55,904 --> 00:06:58,144 happening. And the the margins are so poor 189 00:06:58,144 --> 00:07:00,165 or negative for many of us. 190 00:07:00,544 --> 00:07:01,044 And 191 00:07:01,589 --> 00:07:04,230 there is so many so much solutions out 192 00:07:04,230 --> 00:07:06,550 there, many find and click solution for each 193 00:07:06,550 --> 00:07:08,330 different disease. There's a different 194 00:07:08,709 --> 00:07:10,170 AI algorithm and multiple 195 00:07:11,110 --> 00:07:13,029 that we don't even know how to start 196 00:07:13,029 --> 00:07:15,415 and really execute in a meaningful manner. So 197 00:07:15,415 --> 00:07:17,194 I see this problem of plenty 198 00:07:18,134 --> 00:07:20,314 and both a challenge and an opportunity. 199 00:07:21,095 --> 00:07:23,254 One of the things we are augmenting that 200 00:07:23,254 --> 00:07:25,115 and not solving it in the right manner 201 00:07:25,574 --> 00:07:27,995 is that we are doing our entire governance 202 00:07:28,529 --> 00:07:30,629 to AI adoption roadmap, to strategy, 203 00:07:31,170 --> 00:07:32,629 to implementation science, 204 00:07:33,009 --> 00:07:34,790 to impact measurement and monitoring 205 00:07:35,569 --> 00:07:37,990 separately. Each health system is doing it separately. 206 00:07:38,610 --> 00:07:39,589 And that means 207 00:07:40,414 --> 00:07:42,435 we are duplicating each other work. 208 00:07:43,214 --> 00:07:46,274 And in time of precious resources, that's exactly 209 00:07:46,334 --> 00:07:48,414 opposite of what we should be doing. So 210 00:07:48,414 --> 00:07:50,354 there's a there's a big problem to solve, 211 00:07:50,654 --> 00:07:52,839 and we are trying to solve alone in 212 00:07:52,839 --> 00:07:53,660 each organization, 213 00:07:54,600 --> 00:07:56,379 why we should be solving this together 214 00:07:56,839 --> 00:07:58,139 as a as a scientific 215 00:07:58,919 --> 00:07:59,419 and, 216 00:07:59,879 --> 00:08:00,379 implementation 217 00:08:00,839 --> 00:08:03,240 community. Yeah. And that's exactly what led to 218 00:08:03,240 --> 00:08:04,620 ValleyDI being formed. 219 00:08:05,245 --> 00:08:06,685 And now we are trying to take it 220 00:08:06,685 --> 00:08:08,845 to the next level on execution at scale. 221 00:08:08,845 --> 00:08:09,345 Yeah. 222 00:08:09,884 --> 00:08:12,125 That's fascinating to hear. You know? As you're 223 00:08:12,125 --> 00:08:14,204 getting to that next level, as you're finding, 224 00:08:14,524 --> 00:08:16,925 ways that you you can scale the projects 225 00:08:16,925 --> 00:08:18,605 that work and and then really have that 226 00:08:18,605 --> 00:08:19,105 strong 227 00:08:19,750 --> 00:08:22,149 bend towards execution. How do you make sure 228 00:08:22,149 --> 00:08:23,750 your culture is ready for that? The people 229 00:08:23,750 --> 00:08:25,529 within the organizations are, 230 00:08:25,990 --> 00:08:28,310 supported and educated in the right way so 231 00:08:28,310 --> 00:08:30,569 that this transition is as seamless as possible. 232 00:08:30,629 --> 00:08:31,990 I know it's never easy, but, 233 00:08:32,585 --> 00:08:34,664 how how can you set yourself up for 234 00:08:34,664 --> 00:08:37,865 the best type of implementation or transition, for 235 00:08:37,865 --> 00:08:38,924 new digital technologies? 236 00:08:39,945 --> 00:08:41,705 Yeah. This is such a great question. And 237 00:08:41,705 --> 00:08:43,144 and kind of the route started when I 238 00:08:43,144 --> 00:08:45,465 was at Sana and really strengthened at UC 239 00:08:45,465 --> 00:08:45,965 Davis. 240 00:08:46,549 --> 00:08:49,029 We started a champions group, innovation champions group, 241 00:08:49,029 --> 00:08:50,549 and I was at Mount Sinai. And we 242 00:08:50,549 --> 00:08:51,449 did the formal, 243 00:08:51,990 --> 00:08:54,569 leadership program called digital Davis leadership program. 244 00:08:54,870 --> 00:08:56,169 So I went to the chairs, 245 00:08:56,789 --> 00:08:58,169 through the dean's office, 246 00:08:58,725 --> 00:09:01,445 and asked them, hey. Identify the young faculty 247 00:09:01,445 --> 00:09:03,605 member who's interested in this space. We got 248 00:09:03,605 --> 00:09:07,524 20 faculty members across 20 departments from cancer 249 00:09:07,524 --> 00:09:09,845 to heart failure. And and we trained them 250 00:09:09,845 --> 00:09:12,269 over a period of a year, every two 251 00:09:12,269 --> 00:09:15,149 weeks, very hands on, experientially based, project based 252 00:09:15,149 --> 00:09:16,529 learning, gave them the tools. 253 00:09:17,309 --> 00:09:18,610 And that had been amazing 254 00:09:19,470 --> 00:09:21,629 to see the impact. Now they're five, ten 255 00:09:21,629 --> 00:09:23,970 grants out of it, 10 abstracts, 256 00:09:25,184 --> 00:09:26,565 you know, around 12 papers. 257 00:09:27,264 --> 00:09:29,825 And many of them those people have become 258 00:09:29,825 --> 00:09:31,445 thought leaders within their own specialty 259 00:09:32,625 --> 00:09:34,165 in that. So I do believe, 260 00:09:34,705 --> 00:09:36,325 there's a very big need. 261 00:09:36,699 --> 00:09:38,459 In fact, we need to prime the pump 262 00:09:38,459 --> 00:09:39,759 before we bring technology 263 00:09:40,379 --> 00:09:41,679 because there's a lot of misunderstanding 264 00:09:42,059 --> 00:09:44,139 around that as well. So I think part 265 00:09:44,139 --> 00:09:45,980 of it is we need to do a 266 00:09:45,980 --> 00:09:47,440 double work of 267 00:09:48,139 --> 00:09:50,159 some basic education for everyone. 268 00:09:50,754 --> 00:09:52,855 And and then we also need to have 269 00:09:53,074 --> 00:09:53,574 targeted 270 00:09:53,954 --> 00:09:56,034 deeper education for people who want to go 271 00:09:56,034 --> 00:09:57,735 much more advanced in that regard. 272 00:09:58,194 --> 00:09:59,334 So that's a enormous, 273 00:09:59,954 --> 00:10:00,694 kind of 274 00:10:01,394 --> 00:10:03,174 a need that we we feel. 275 00:10:03,500 --> 00:10:05,980 And we started doing executive education at Valley 276 00:10:05,980 --> 00:10:07,740 Day, which had been super successful. We meet 277 00:10:07,740 --> 00:10:10,059 biweekly and share and learn between each other 278 00:10:10,059 --> 00:10:10,879 the best practices. 279 00:10:11,339 --> 00:10:12,559 But I believe we need, 280 00:10:13,019 --> 00:10:15,279 a different layer completely to take an enterprise 281 00:10:15,339 --> 00:10:16,159 wide approach. 282 00:10:16,565 --> 00:10:18,084 And we are kind of in early stage 283 00:10:18,084 --> 00:10:19,704 of thinking how do we bring this 284 00:10:20,084 --> 00:10:21,544 Gen AI and AI Academy, 285 00:10:22,404 --> 00:10:24,345 to really train people at masses, 286 00:10:25,125 --> 00:10:27,945 but not in a kind of boring compliance 287 00:10:28,084 --> 00:10:30,830 based manner than what we see on the 288 00:10:31,210 --> 00:10:33,230 the portals or learning management system, 289 00:10:34,009 --> 00:10:37,629 but in a transformative manner, expediential manner, empowerment 290 00:10:37,690 --> 00:10:40,730 manner. And that may mean a combination of 291 00:10:40,730 --> 00:10:41,389 in person 292 00:10:41,894 --> 00:10:42,715 launch days, 293 00:10:43,654 --> 00:10:44,955 vision days to, 294 00:10:45,575 --> 00:10:46,955 to project based learning, 295 00:10:47,654 --> 00:10:49,654 with some synthetic data or something to play 296 00:10:49,654 --> 00:10:52,375 with. So we call it everyone gets benefit, 297 00:10:52,375 --> 00:10:54,855 and some people can really take it much 298 00:10:54,855 --> 00:10:56,519 more deeper if they have to. So it's 299 00:10:56,519 --> 00:10:57,899 precision based as well. 300 00:10:58,919 --> 00:11:00,679 Got it. That's helpful to know. Thank you 301 00:11:00,679 --> 00:11:02,919 for walking us through that. Now looking ahead 302 00:11:02,919 --> 00:11:05,639 and into the future, how do you see, 303 00:11:05,959 --> 00:11:08,279 thinking about growth and developing and and continuing 304 00:11:08,279 --> 00:11:09,179 to add value, 305 00:11:09,684 --> 00:11:11,605 to hospitals and health systems as well as 306 00:11:11,605 --> 00:11:14,324 the physicians who are, you know, really interested 307 00:11:14,324 --> 00:11:16,964 in learning more and digging deeper into the 308 00:11:16,964 --> 00:11:19,304 AI technology driven application space? 309 00:11:20,245 --> 00:11:23,269 Yeah. I think one thing is clear. We 310 00:11:23,269 --> 00:11:25,210 have a affordability crisis here. 311 00:11:25,590 --> 00:11:27,990 What is being offered even if we want 312 00:11:27,990 --> 00:11:29,450 it, then we just can't. 313 00:11:29,910 --> 00:11:31,669 So there's a very famous quote that the 314 00:11:31,669 --> 00:11:33,750 AI technologies are becoming cheaper, but the people 315 00:11:33,750 --> 00:11:35,509 to implement that are becoming more and more 316 00:11:35,509 --> 00:11:36,730 expensive or not available. 317 00:11:37,065 --> 00:11:39,245 So we did some kind of internal 318 00:11:39,785 --> 00:11:40,285 assessment, 319 00:11:40,665 --> 00:11:42,684 from the members, and what we found was 320 00:11:42,985 --> 00:11:45,644 to really bring in a very trustworthy, responsible, 321 00:11:46,184 --> 00:11:49,465 secured manner and to build a team that 322 00:11:49,465 --> 00:11:51,565 understands that how to bring it and monitor. 323 00:11:52,089 --> 00:11:54,490 That's an effort of 5 to $10,000,000 324 00:11:54,490 --> 00:11:55,870 a year, at least to start. 325 00:11:56,490 --> 00:11:59,289 Right? Because you for example, like UC Davis 326 00:11:59,289 --> 00:12:01,690 has health has 800 applications, IT. Four hundred 327 00:12:01,690 --> 00:12:03,470 of them will have some kind of automation 328 00:12:03,529 --> 00:12:04,750 AI technology component 329 00:12:05,289 --> 00:12:07,309 to for me to validate all of them, 330 00:12:07,504 --> 00:12:08,884 assess their equity play, 331 00:12:09,345 --> 00:12:10,565 see if they don't have biases, 332 00:12:11,024 --> 00:12:13,684 and, they have repeatable replicable in their recommendations, 333 00:12:14,225 --> 00:12:16,304 that's a huge amount of burden I'm adding 334 00:12:16,304 --> 00:12:17,445 in addition to cybersecurity. 335 00:12:18,865 --> 00:12:21,184 So so we are finding the the really, 336 00:12:21,184 --> 00:12:23,330 the the elephant in the room is affordability 337 00:12:23,470 --> 00:12:25,090 crisis for us to transform. 338 00:12:25,950 --> 00:12:27,549 Now now the good thing is the other 339 00:12:27,549 --> 00:12:28,529 part of that is 340 00:12:29,389 --> 00:12:31,629 the use cases have to be brought, so 341 00:12:31,629 --> 00:12:33,309 the impact has to be measured as well. 342 00:12:33,309 --> 00:12:35,084 So a lot of our effort, and that's 343 00:12:35,084 --> 00:12:36,445 why we are kind of shaping what we're 344 00:12:36,445 --> 00:12:39,004 calling the transformation collective to the next stage 345 00:12:39,004 --> 00:12:39,904 of valid AI, 346 00:12:40,764 --> 00:12:43,105 where multiple organizations with valid 347 00:12:43,565 --> 00:12:45,565 AI who are ready on a maturity level 348 00:12:45,565 --> 00:12:47,105 and have the right leadership, 349 00:12:47,725 --> 00:12:48,304 in place 350 00:12:48,769 --> 00:12:50,389 to take the next stage of transformation 351 00:12:51,250 --> 00:12:51,750 where 352 00:12:52,289 --> 00:12:54,230 we can find the best use cases, 353 00:12:55,009 --> 00:12:56,690 but we are also able to, through share 354 00:12:56,690 --> 00:12:57,190 team, 355 00:12:57,570 --> 00:13:00,389 decrease the cost. And and really kind of 356 00:13:00,690 --> 00:13:02,975 no cost comes we actually look at the 357 00:13:02,975 --> 00:13:05,235 impact first, problem first approach 358 00:13:05,774 --> 00:13:07,375 rather than what's happening in the market is 359 00:13:07,375 --> 00:13:09,375 a problem of plenty and solution based approach. 360 00:13:09,375 --> 00:13:11,294 Oh, we have so many solutions. Which one 361 00:13:11,294 --> 00:13:13,154 I want as a point and click solution? 362 00:13:13,375 --> 00:13:15,455 Worse if these are the five strategic problems 363 00:13:15,455 --> 00:13:18,070 we have in our health system, Let's work 364 00:13:18,070 --> 00:13:19,850 backwards to say, fine. 365 00:13:20,389 --> 00:13:22,789 Maybe a singular platform can solve many of 366 00:13:22,789 --> 00:13:25,750 those issues and create economies of scale, economies 367 00:13:25,750 --> 00:13:26,409 of affordability, 368 00:13:27,110 --> 00:13:29,669 and economies for impact. So I think it's 369 00:13:29,669 --> 00:13:31,190 a I see that as a very big 370 00:13:31,190 --> 00:13:31,690 opportunity 371 00:13:32,070 --> 00:13:34,325 to view the lens of problem based learning 372 00:13:34,325 --> 00:13:34,825 here. 373 00:13:35,605 --> 00:13:37,605 Got it. That's super helpful to know. And 374 00:13:37,605 --> 00:13:40,085 and, you know, really cool to have that 375 00:13:40,085 --> 00:13:42,245 opportunity now to move on to the next 376 00:13:42,245 --> 00:13:45,044 level and and, really integrate into the health 377 00:13:45,044 --> 00:13:45,544 system. 378 00:13:46,379 --> 00:13:48,779 As you're looking at growth development, and I 379 00:13:48,779 --> 00:13:50,779 know technology is moving quickly in the health 380 00:13:50,779 --> 00:13:52,860 care space, what are still some of those 381 00:13:52,860 --> 00:13:55,120 risks and challenges that you see, 382 00:13:55,740 --> 00:13:57,980 and and what investment will it take in 383 00:13:57,980 --> 00:13:59,759 order to, really realize 384 00:14:00,125 --> 00:14:01,024 the overall benefit, 385 00:14:01,404 --> 00:14:03,184 for some of these technologies as well? 386 00:14:03,804 --> 00:14:05,804 Yeah. I think the the biggest risk we 387 00:14:05,804 --> 00:14:09,024 see is the the risk of bringing technology, 388 00:14:09,485 --> 00:14:11,024 which is not along with strategy. 389 00:14:11,485 --> 00:14:13,084 Spending money into it and the money drives 390 00:14:13,084 --> 00:14:14,784 away, so you can't really make an impact. 391 00:14:15,200 --> 00:14:17,839 And, outdoing it in a manner, that actually 392 00:14:17,839 --> 00:14:20,000 opens up cybersecurity risk or other kind of 393 00:14:20,000 --> 00:14:22,159 risk and make your, things. So I think 394 00:14:22,159 --> 00:14:23,600 those are the risks we see are very 395 00:14:23,600 --> 00:14:24,100 tangible. 396 00:14:24,559 --> 00:14:26,500 The main gap we see, 397 00:14:27,279 --> 00:14:28,820 is the talent gap, 398 00:14:29,644 --> 00:14:32,764 to really allow us from and we published 399 00:14:32,764 --> 00:14:34,925 a paper, you know, five or six of 400 00:14:34,925 --> 00:14:36,785 us, from different academic centers. 401 00:14:37,245 --> 00:14:39,884 UC San Diego, UC Davis, we combined, and 402 00:14:39,884 --> 00:14:42,850 we talked about how we need to bring 403 00:14:43,149 --> 00:14:44,930 AI implementation science centers. 404 00:14:45,790 --> 00:14:47,950 The variability most of people talk about model 405 00:14:47,950 --> 00:14:48,450 variability, 406 00:14:48,910 --> 00:14:51,490 but we truly believe the variability is more 407 00:14:51,710 --> 00:14:53,330 in the real world even, 408 00:14:53,790 --> 00:14:56,350 because that's when we reduce the models or 409 00:14:56,350 --> 00:14:57,554 AI models to practice. 410 00:14:58,674 --> 00:15:00,534 Who touches them, how they touch, 411 00:15:01,075 --> 00:15:03,075 and what does the impact happens, and that 412 00:15:03,075 --> 00:15:05,174 does lead to a positive outcome or not. 413 00:15:05,475 --> 00:15:07,414 There's so much variability in that workflow 414 00:15:07,875 --> 00:15:10,214 that 80%, sometimes, we call it as variability 415 00:15:10,274 --> 00:15:12,600 at the implementation science level and not at 416 00:15:12,600 --> 00:15:15,240 the model level. While everyone talks about model 417 00:15:15,240 --> 00:15:16,299 validation, we believe 418 00:15:16,679 --> 00:15:17,740 to really derisk 419 00:15:18,120 --> 00:15:20,620 the AI adoption, we need a strategic lens. 420 00:15:20,839 --> 00:15:22,600 We need to address the talent gap through 421 00:15:22,600 --> 00:15:23,580 a shared service, 422 00:15:24,279 --> 00:15:26,725 and we need And that by addressing that 423 00:15:26,725 --> 00:15:29,365 both strategic lens and the right talent can 424 00:15:29,365 --> 00:15:29,865 help 425 00:15:30,325 --> 00:15:32,424 create AI implementation signs of excellence. 426 00:15:33,284 --> 00:15:34,804 And then we can have a network of 427 00:15:34,804 --> 00:15:36,904 those AI implementation signs of excellence, 428 00:15:37,284 --> 00:15:37,784 which 429 00:15:38,190 --> 00:15:40,590 support each other and complement each other and 430 00:15:40,590 --> 00:15:41,649 complete each other, 431 00:15:42,029 --> 00:15:44,910 rather than duplicate each other work again, and 432 00:15:44,910 --> 00:15:47,009 that doesn't lead to the ecosystem evolving. 433 00:15:47,629 --> 00:15:50,210 So I think, it's just a fascinating time, 434 00:15:50,750 --> 00:15:52,769 where the opportunity is huge 435 00:15:53,245 --> 00:15:53,985 to impact 436 00:15:54,764 --> 00:15:55,264 clinicians, 437 00:15:55,644 --> 00:15:57,024 like, through AI scribes, 438 00:15:58,045 --> 00:16:00,945 through patients, like, through digital avatar and automation 439 00:16:01,004 --> 00:16:03,165 and having a post call for them which 440 00:16:03,165 --> 00:16:06,304 knows them automatically guides them through conversation layer 441 00:16:06,605 --> 00:16:07,904 through operational efficiency. 442 00:16:08,420 --> 00:16:09,940 So a lot of the things through agentic 443 00:16:09,940 --> 00:16:12,019 workflows can be done automatically. We do not 444 00:16:12,019 --> 00:16:14,580 have to repeat and manual stuff and really 445 00:16:14,580 --> 00:16:18,040 liberate the data to intelligence layer, because unstructured 446 00:16:18,100 --> 00:16:20,180 data, which is 80% of health system data, 447 00:16:20,180 --> 00:16:22,259 can now be liberated to intelligence layer of 448 00:16:22,259 --> 00:16:23,240 generating AI. 449 00:16:24,514 --> 00:16:26,915 Absolutely. Wow. That's, really cool to hear and 450 00:16:26,915 --> 00:16:29,955 definitely amazing to see how the different health 451 00:16:29,955 --> 00:16:32,035 systems are working together, as you mentioned, just 452 00:16:32,035 --> 00:16:33,815 to share information and share, 453 00:16:34,835 --> 00:16:36,434 share how they're doing things. And so, you 454 00:16:36,434 --> 00:16:37,955 know, you can build upon each other's work 455 00:16:37,955 --> 00:16:40,139 and continue to advance in the future. Before 456 00:16:40,139 --> 00:16:42,139 we wrap up here, I'm curious, where do 457 00:16:42,139 --> 00:16:43,519 you see some of the best opportunities 458 00:16:43,899 --> 00:16:45,200 for physicians especially, 459 00:16:45,980 --> 00:16:46,720 in the future? 460 00:16:48,059 --> 00:16:48,960 Yeah. I think, 461 00:16:49,500 --> 00:16:51,340 one is I think we I had predicted 462 00:16:51,340 --> 00:16:53,004 two, three years ago, we're gonna have AI 463 00:16:53,004 --> 00:16:55,324 scribes like water. It should be ubiquitous, and 464 00:16:55,324 --> 00:16:57,024 everyone should be able to use it, and 465 00:16:57,245 --> 00:16:58,924 the prices will come down. And that's what's 466 00:16:58,924 --> 00:17:01,164 happening because the need is so real from 467 00:17:01,164 --> 00:17:03,245 our AI scribes to decrease the burden of 468 00:17:03,245 --> 00:17:06,044 documentation for clinicians. And we're gonna see extension 469 00:17:06,044 --> 00:17:08,704 of that intelligence layer doing more and more, 470 00:17:09,779 --> 00:17:12,099 in terms of intelligence and decisions and putting 471 00:17:12,099 --> 00:17:14,440 orders. So that's just an amazing, 472 00:17:14,899 --> 00:17:15,399 amazing 473 00:17:16,019 --> 00:17:16,519 opportunity. 474 00:17:16,819 --> 00:17:18,179 And this is the first time we have 475 00:17:18,179 --> 00:17:20,819 intelligence layer between a patient and a physician 476 00:17:20,819 --> 00:17:22,039 that's on top of a EMR. 477 00:17:22,544 --> 00:17:24,944 So the magic can happen here. We call 478 00:17:24,944 --> 00:17:25,924 it headless EMR, 479 00:17:27,024 --> 00:17:28,865 that that can potentially shape up. We all 480 00:17:28,865 --> 00:17:31,024 dreamt of it, but this finally can become 481 00:17:31,024 --> 00:17:31,684 a reality. 482 00:17:32,065 --> 00:17:34,404 But I'm equally and and maybe more excited. 483 00:17:34,900 --> 00:17:36,819 There are limited number of clinicians, and we 484 00:17:36,819 --> 00:17:39,779 can empower them, and that's so humongous because 485 00:17:39,940 --> 00:17:41,400 but they also have to serve 486 00:17:42,259 --> 00:17:44,019 millions and billions of patients if you look 487 00:17:44,019 --> 00:17:45,240 at worldwide lens. 488 00:17:45,779 --> 00:17:47,319 How can we empower patients 489 00:17:48,115 --> 00:17:49,954 to be a partner in their own journey? 490 00:17:49,954 --> 00:17:51,394 And that's my lens from a public health 491 00:17:51,394 --> 00:17:52,375 lens as well. 492 00:17:53,315 --> 00:17:54,054 When I 493 00:17:54,355 --> 00:17:55,974 I the words with my patients, 494 00:17:56,835 --> 00:17:58,755 it's not just me telling them what to 495 00:17:58,755 --> 00:18:00,595 do, but also empowering them so they can 496 00:18:00,595 --> 00:18:01,414 learn knowledge. 497 00:18:01,714 --> 00:18:03,255 And that way their behavior, 498 00:18:04,200 --> 00:18:04,700 changes, 499 00:18:05,640 --> 00:18:06,380 and that's 500 00:18:06,759 --> 00:18:08,359 stuff they have to do, and and we 501 00:18:08,359 --> 00:18:10,599 can empower them. But if we can create 502 00:18:10,599 --> 00:18:13,099 digital avatars, and we started doing that in, 503 00:18:13,480 --> 00:18:16,679 wellness space, to the venture studio, we're planning 504 00:18:16,679 --> 00:18:18,954 to bring more of them, then they have 505 00:18:18,954 --> 00:18:21,534 guidance with them 247365. 506 00:18:21,835 --> 00:18:23,855 Right? We can even I give this example 507 00:18:23,914 --> 00:18:26,634 from physician perspective, IVD perspective. If I can 508 00:18:26,634 --> 00:18:28,875 actually take what I tell my patients, maybe 509 00:18:28,875 --> 00:18:30,255 I can take best 10 510 00:18:30,720 --> 00:18:32,799 physicians in inflammatory bowel disease, what they tell 511 00:18:32,799 --> 00:18:34,179 their patients at certain time. 512 00:18:34,480 --> 00:18:37,039 We can make our collective digital avatar, and 513 00:18:37,039 --> 00:18:39,039 that can be with the patient 247, 3 514 00:18:39,039 --> 00:18:41,940 60 5 answering plates in a trusted manner, 515 00:18:42,319 --> 00:18:43,919 but with a layer of human in the 516 00:18:43,919 --> 00:18:44,419 loop, 517 00:18:45,335 --> 00:18:46,555 telling them which are 518 00:18:47,174 --> 00:18:48,714 tested, repeatable, replicable, 519 00:18:49,255 --> 00:18:51,575 but anything that's vague or that comes in 520 00:18:51,575 --> 00:18:53,974 a different point of decision making bubbles up 521 00:18:53,974 --> 00:18:55,115 to the human's level, 522 00:18:55,414 --> 00:18:57,275 we can create a great way of 523 00:18:57,734 --> 00:18:58,714 human led 524 00:18:59,095 --> 00:18:59,994 AI transformation. 525 00:19:01,220 --> 00:19:03,640 Right? And and that will 526 00:19:04,340 --> 00:19:05,960 change complete care 527 00:19:06,980 --> 00:19:08,200 from one to one 528 00:19:08,660 --> 00:19:10,980 where we believe a clinician and a patient 529 00:19:10,980 --> 00:19:12,900 has to be in the same space in 530 00:19:12,900 --> 00:19:13,880 the same time 531 00:19:14,695 --> 00:19:16,394 but truly one too many 532 00:19:16,855 --> 00:19:18,795 where the same clinician can impact 533 00:19:19,095 --> 00:19:22,055 millions of patients at once. And we had 534 00:19:22,055 --> 00:19:23,835 seen that in COVID time 535 00:19:24,215 --> 00:19:26,234 and earlier part with the rule based AI. 536 00:19:26,455 --> 00:19:27,974 We were able to build a bot within 537 00:19:27,974 --> 00:19:30,720 eight days and touch 1,000,000 patients through EMR 538 00:19:30,720 --> 00:19:32,240 and 55,000 539 00:19:32,240 --> 00:19:33,460 patients were being monitored. 540 00:19:34,079 --> 00:19:35,940 In the age of generative AI, 541 00:19:36,319 --> 00:19:38,400 we can have a hundred x, 2,000,000 x, 542 00:19:38,799 --> 00:19:39,920 that kind of, 543 00:19:40,480 --> 00:19:42,319 value stream if we do it the right 544 00:19:42,319 --> 00:19:42,819 way. 545 00:19:43,154 --> 00:19:45,575 So so super excited about that possibility, 546 00:19:47,075 --> 00:19:48,134 of working together, 547 00:19:48,595 --> 00:19:50,615 with with many of the partners here, 548 00:19:51,315 --> 00:19:52,775 to to march towards that. 549 00:19:53,315 --> 00:19:55,428 That's excellent to hear. Doctor, Otrincha, thank you 550 00:19:55,428 --> 00:19:56,868 so much for joining us on the podcast 551 00:19:56,868 --> 00:19:58,868 today. This has been a wonderful conversation, and 552 00:19:58,868 --> 00:20:00,308 I look forward to connecting with you again 553 00:20:00,308 --> 00:20:00,808 soon. 554 00:20:01,428 --> 00:20:02,248 Great. Thank