 Live from Boston, it's theCUBE, covering IBM Chief Data Officer Summit. Brought to you by IBM. Welcome back everyone to theCUBE's live coverage of the IBM CDO Summit here in Boston, Massachusetts. I'm your host, Rebecca Knight, along with my co-host, Paul Gillan. We're joined by Dr. Prakati Suthwar Santikari, known as Dr. Santi. He is the Vice President and Global Chief Data Officer at e-research technology. Thank you so much for coming back on theCUBE. Yeah, thank you for inviting me. So, Dr. Santi, tell our viewers a little bit about e-research technology. You're based in Marlboro. Yeah, so we are in Boston, but ERT has been around since 1977, and we are a data and technology company that minimizes the risks and uncertainties within clinical trial space. And our customers are pharmaceutical companies, biotechnology companies, medical device companies, and they really trust us in terms of running their clinical trials on our platform. So, we have been around over 40 years, so we have seen a thing or two in this space. It's a very complex domain, and very highly regulated, as you know, because it's dealing with patients' lives. So, we take huge pride in what we do. We know how involved clinical trials can be long, very expensive. How are the new tools of big data impacting that cost? Well, that has been an age-old problem within the clinical trials. Usually, a drug takes about eight to 12 years, and costs about $2 billion from start to commercialization. So, it's a very lengthy, manual, and arduous process. So, there are lots going on in this clinical trial domain that tries to shorten the timeline, and employing of big data technologies, modern data platform to expedite data processing, data collection from mobile devices and health technologies and all this. Artificial intelligence is playing a big role in terms of disrupting some of these domains, particularly if you see the protocol development down to patient selection, down to study design, then study monitoring. So, you need to do all those things, and each takes a long, long, long time. So, AI, with the big data technologies, is they're really making a difference. In what ways? For example, patient selection is one of the huge pinpoints in any clinical trial, because without patients, there are no clinical trials. Particularly, when you try to launch a drug, you'll have to identify the patients, select the patients, and not only select the patients, you have to make sure those patients stay with the clinical trials throughout the duration of the trial. So, patient engagement is also a big deal. So, with these big data technologies, like now you can see all these mobile health devices that the patients are wearing, using which you can monitor them. You can send them a reminder, take your drug, or you can send a text saying that there will be a clinical visit at that site, come at seven o'clock, don't come at nine o'clock. So, these kind of encouragement and constant feedback loop is really helping patients stay engaged, and that is critical. Then, matching patients with a given clinical trials is a very manual and arduous process. So, that's where algorithms is helping. So, they're just cranking up real-world evidence data, for example, claims data, prescription data, and other type of genomic data, and they're matching patients and the clinical trial needs. Instead of just fishing around in a big pond and find out, okay, I need three patients, so go and fish around the world to get those three patients. That's why current process is very manual, and these AI techniques and BI technologies are really disrupting this industry. So, are the pharmaceutical companies finding that clinical trials are better today? Because patients are more engaged, and they are getting, as you said, this constant reminder, take your drug, stay with us. Do you think that they are, in fact, giving them better insights into the efficacy of the drug? Yes, because so that you will see their compliance rate is increasing. So, because remember, when they have to fill out all these diaries, like morning diaries, evening diaries, when they are taking which medicine, when they are not taking, it used to be all manual, paper-driven. So, they would forget, and particularly think about a terminally ill patient. Each day is so critical for them. So, they don't have patients, nor do they have time to really maintain a manual diary. Nor do their caregivers have the time, right? So, this kind of automation is really helping. And that is also encouraging them as well, that, yeah, somebody is really caring about it. We are not just a number. Patient is not a number. Somebody is really relating to them. So, patient engagement, we have a product that specifically focuses around patient engagement. So, we do all these phase one through phase four trials, one, two, three, four, and then post-marketing, obviously. But through the entire process, we also do patient engagement, so that we help our customers, like pharmaceutical companies and biotechnology companies, so that they can run their trials with confidence. How about analyzing the data that you collect from the trials? Are you using new techniques to gain insights more quickly? Yes, we are. We just recently launched a modern data platform, Data Lake, where we are consolidating all the data, anonymizing it, and then really applying AI techniques on top of it. And also, it is giving us real-time information for study monitoring, like which site is not compliant, which patients are not compliant. So, the data quality is a big deal in clinical trials, because if the quality is good, then FDA approval, you know, there's a chance that FDA may approve. But if the data quality is bad, forget about it. So that's why I think the quality of the data and monitoring of that trial real-time to minimize any risks before they become risks. So you have to be preemptive. So that's why these predictive algorithms are really helping, so that you can monitor the site, you can monitor individual patients through M-Health devices and all this, and really pinpoint that, hey, your clinical trials are not going to, you know, end on time, nor on budget, because see the actual situation here, so do something instead of waiting 10 years to find that out. So huge cost-saving and efficiency gain. I want to ask about data in healthcare in general, because one of the big tensions that we've talked about today is sort of what the data is saying versus what people's gut is saying. And then, you know, in industry, it's the business person's gut, but in healthcare, it's the doctor, the caregiver's gut. So how have you seen data, or how has data perceived, and is that changing in terms of what the data shows that the physician about the patient's condition and what the patient needs right then and there versus what the doctor's gut is telling him that the patient needs? Yeah, and that's what that augmentation and complementary nature, right? So AI and doctors, they're like complimenting each other. So predictive algorithm is not replacing doctors, you know, the expertise. So you still need that. What AI and predictive algorithm is playing a big role is in expediting that process. So instead of sifting through manual, you know, the documents or sifting through this much amount of document, they would only need to do this much of document. So that way, it's minimizing that time horizon. It's all about efficiency again. So AI is not going to be replacing doctors anytime soon. We still need doctors because remember, a site is run by a primary investigator and primary investigator owns that site. That's the doctor. That's not a machine. That's not an AI algorithm. So his or her approval is the final approval. But it's all about efficiency, cost-cutting and bringing the drugs to the market faster. If you can cut down these 12 years by half, think about that. Not only are you saving lots of money, you are also helping patients because those drugs are going to get to the market six years earlier. So you are saving lots of patients in that regard as well. One thing that technologies like Watson can do is sort through, read millions of documents, lab reports and medical journals and derive insights from them. Is that helping in the process of perhaps avoiding some clinical trials or anticipating outputs earlier? Yes, because if you see Watson run a clinical study with Cleveland Clinic recently or Mayo Clinic, I think, or maybe both, where they reduced the patient recruitment time by 80%. 80% because they sweep through all those documents, EMR results, claims data, all this data they combine. Filter down, filter down, and then say for this clinical trial, here are the 10 patients you need. It's not going to recommend who those 10 patients are, but it will just tell you that the average location is that, so that you just focus on getting those 10 patients quickly instead of wasting nine months to research on the 10 patients. And that's a huge, huge deal. And how can you trust that that is right? You know, I think that's another question that we have here, it's a big challenge. It is a challenge, it is a challenge. Because AI is all about maths and algorithms, right? So when you, so it's like an input black box, output. So that output may be more accurate than what you perceive it to be. But that black box is what is tripping me up here. So what is happening is sometimes, oftentimes if it is a deep learning techniques or that kind of lower level AI techniques, it's very hard to interpret that results. So people are, people will keep coming back to you and say, how did you arrive at that results? And that's where most of the, there are techniques, like machine learning techniques that are easily interpretable. So you can convince FDA folks or other folks that here is how you got to it. But there are deep learning techniques that Watson uses, for example. People will come and, you know, how did you arrive at that? And it's very hard because those neural networks are multi-layered and all about maths. But as I said, output may be way more accurate, but it's very hard to decipher. Right, exactly. So that's a challenge. So that's a trust issue in that regard. Right, right. Well, Dr. Santi, thank you so much for coming on theCUBE. It was great talking to you. Okay, thank you very much. Thanks for inviting me. I'm Rebecca Knight for Paul Gillan. We will have more from the IBM CDO Summit in just a little bit.