 Live from Las Vegas, it's theCUBE. Covering Informatica World 2019. Brought to you by Informatica. Welcome back everyone to theCUBE's live coverage of Informatica World 2019 here in Sin City, Nevada. I'm your host, Rebecca Knight, along with my co-host, John Furrier. We have two guests for this segment. We have Pinkrose Hamilton. She is the VP Business Intelligence at Hackensack Meridian Health. Thanks for coming on the show. Thank you for having me. And we have Andy Craigow. He is a managing consultant at InfoVarity. Thanks so much, Andy. Thanks for having me. So, tell us a little bit about this partnership between Hackensack and InfoVarity. Well, we were looking for an implementation partner. We were looking for the skills to come in and help us really implement MDMs specifically. We're also implementing a few other technologies that we could probably speak about, but that's how we got connected. So, tell us a little bit about what life was like before MDM. What were sort of the obstacles, the challenges that you were wrestling with? So, Hackensack Meridian Health is the largest health system in New Jersey. And we are a very fast growing, we like to consider ourselves disruptive health industry in New Jersey. And so, because of that, we were growing and acquiring mergers acquisitions and many different EMRs, many different physician credentialing systems were involved in this. So, we had to make a decision of, do we wait till we're all on one system, which we all know will never happen, or never happen in time sometimes. So, we decided to do the MDM approach, which made the most sense to us. One of the things that's interesting, we talk, we go to hundreds of events, we go to talk to a lot of experts and practitioners. And everyone buys into cloud at some level. If you're cloud native startup, certainly born in the cloud, great benefits. Data's critical, because in SaaS, data's great if you have it, because you can feed machine learning, you can take more risk, be agile, and more risk, more reward. And the apps, it's all good, right? On the enterprise side, on premises, legacy kind of kicks in. If data can't feed machine learning or can't feed the app, AI really can't be enabled. This becomes a key challenge in the industry. How do you guys look at that? Because as you lay out, it's not a simple answer, go to the cloud, just do on-prem and you got to think about an architecture. What are you guys doing with the data? Where the data is stored? How do you think about it? What's some advice and best practice can you share? Well, I consider data storage being more like a house you're living in, right? So we buy our starter homes and we start our families. And then we outgrow this house. And then we have to say, okay, I need a bigger house and we start growing. And so data's run pretty much the same way. We start outgrowing our on-prem houses. And so now we're moving out and we're moving to bigger and better things, which is the cloud. And so I think hybrid is where we start, right? We can't start with, okay, everybody move out and move into this new house. It's let's go build this new house somewhere else. Let's test it out and see if we like it. So that's my thought process around data. So you got the addition. That's got to work with all the plumbing, right? So it's the same thing. And then you got more track homes and you got electronic cars that go in between automation. So this is more of a systems view. Take care of the operational piece. Absolutely. Then think about developer angle. What's the, how does that architecture look? So in terms of what we're trying to do right now, I mean, it has to be kind of short-term vision with kind of a larger scale architecture. So, as Pink was saying, in terms of the hybrid architecture, if we are able to develop reusable cleanse functions, such as kind of the address doctor functionality, we're reaching out to a third party service, bringing in more enriched information. We have that in a non-prem model right now, but in the future, that configuration and work will easily transition into that cloud architecture. So we're trying to keep our eye on the future and make sure that things are reusable as we move forward. And how do you two work together? I mean, this is such an interest, I mean, in this age of co-opetition, and you're not necessarily competitors, of course, but how do you work together to come up with the right solutions? I mean, what does that look like, the partnership? Well, we totally hate each other. That's right, yeah. This is the first we've talked in a while, yeah. No, the partnership, I think, we hit it off right from the beginning. It was just a matter of, when we acquire new technologies and that decision of how much time and effort is it going to take for me to train my team and to identify the right folks on my team and what work am I going to take away from them in order to give them this additional work and this learning curve that needs to go into place? So I think we have to augment our teams with experts like Infraverity to come in and say, this is how this tool functions and sometimes we bring in the technologies and we kind of just crack it open but we don't really get the full use of it to understand exactly every bell and whistle we can take advantage of. And these guys are the experts that help us do that. And it's always a challenge. I mean, I think data's been center of the Vapors for many, many years. It's kind of mainstream now and you can't look with the headlines these days without hearing for one year anniversary of GDPR privacy. So there's always been that risk management, compliance stuff that's been around. Certainly you guys know that but every day there's a new thing. Oh, if you've got cloud, you've got geo regions here in this country, you've got in this, that country. So as more regulatory kind of things kind of creep over who knows, maybe blockchains out there. So again, all these things are circling around complexity which constrains data, not necessarily freeze it so much. Well, maybe build software. So how does Informatica and customer deal with this because I'd imagine you'd have to build an abstraction layer, have to be some tooling around it, monitoring. Yeah, what's your take on this complexity? So in terms of like an architecture perspective, you know, we consolidated all of the different kind of silos of patient data into a centralized repository. Historically, you know, you would build kind of a lot of point-to-point feeds based on a certain application. We'd build some custom work and we would ship them off some data. But really what we want to do is be able to master once and publish to a canonical model that's more of self-service and hub and spoke. So as consumers and customers of the data need to come and get it, they can come to a centralized place. We can augment what data is available there and kind of scale that with the architecture across real-time capabilities, cloud and other use cases that we come across. Do you feel good, data's frictionless, it's out there, it's addressable? In terms of the vision that we're on, so I mean it's a couple steps at a time, but in terms of the journey and the set of tools that we have, that's definitely where we're going, so. I want to ask you about the skills gap. One of the things that has emerged is that in the healthcare industry, it is much more evolved in the sense of that they, there's an understanding of how to work with data. And perhaps because you've just always worked with more data than say a retail company or a consumer product company. So first of all, how big a problem is this for hack and sack variety in health? Is it as bad as the headlines suggest? And also what are you doing to combat it? So our main goal is to take care of the patient, right? And so when a patient is introduced to our system, we want to be able to take care of that patient and their family members in the best possible way that we can. So if we're working with a very disparate organization where we're on multiple EMRs specifically, it's hard for us to identify that episode of care for that patient. So the MDM piece particularly with the patient domain allows us to do that. It allows us to view the entire episode of care for that patient to see, you know, you went to these doctor's offices, you had these things done, you went to this lab, you had these tests done, you went to the hospital, you had this procedure, and this is what your follow-up looked like. And we're also conscious of the patient's expense in all of this, as well as what's the provider's expense, what is the payer's expense. So you want to make it cost effective, you want to make it accessible, so that are there services that a certain zip code or patient population needs that we're not providing that we can provide? And so this is the whole entire continuity of care to take care of our patients the best way we can. My daughter just graduated college this week in Cal, first ever data analysis college class, an inaugural class, so shows how early it is, you know, Cal's great school been doing data for a while. Data is a huge opportunity, whether it's women in tech, new service area comes up, whether you don't need to be a hardcore programmer to get into the data business, but there's certain patterns we're seeing emerge that you don't have to have a certain degree because the jobs that are open, there's no degree for it. There's only the first class that's graduated from Berkeley. So I got to ask you for folks in high school or parents out there or anyone looking to re-skill, what specific foundational and or advanced skill sets should people be looking at? If they really want to get into data, and it could be anything, so what, love to get your take on what you think those skills are for people out there that they want to learn something new and ride the wave. So I'll start a little bit. I think a lot of people get really technical with data, but I think you really have to understand data within business context. I mean, if you're looking at a physician record, understanding the type of physician, maybe where the care was administered, you have to really think about, okay, what am I trying to solve? What pain point am I looking at? So it's not about relational databases and writing SQL. You really have to understand the functional purpose of data within the business problem that you're seeing. Well, sort of machine learning's hot. The nerds go there, the geeks go there, but there's a bigger picture than just coding and... There's a whole data strategy that you need to consider and kind of plug and play as you go along and really understanding the data within the business context is key. I'm so glad you asked that question because I'm going to give a different viewpoint from this. I have a daughter who's a junior in high school and she's preparing her career path and so she wants to follow mom's career path and wants to do data science. So it's very exciting for me. I'm actually a role model, which you don't never expect your children to think of you as much. Congratulations. Yeah, so she picked up a few SQL classes early on in high school. I think the underlining foundation of coding is probably a little bit important to get that piece of it because when you're leading the function and definitely knowing the business knowledge, when we start any project, we go in and we start with discovery, right? What is it that you do? How do you do it? What are your workflows? What do they look like? So that's definitely key, but adding in that technical piece makes you that perfect data science human that I would look for as an employer. It's certainly evolving. There's no one yet played because there's so many diverse opportunities to dig in from visualization to ethics, coding to business value, unbelievable. Yeah, great. Well, Pink and Andy, thank you both so much for coming on the show. Thank you so much for having me. Lots of great advice for newly minted graduates. That's right. Thank you. I'm Rebecca Knight for John Furrier. You are watching theCUBE.