Description: The article below is mainly intended to center on the point of maurices capital one. The content will be related to cloud analytic summit and fireside chat with Capital One which started with going to agile and moving to a developer environment that was continuous integration and continuous delivery.
Cabin one has gone through this major transformation over the last a couple of years, tell me about something about that, it’s been an amazing journey, I think for the company to take this steps as saying we are going to take our applications in our entire ecosystem and go cloud native, put everything in the cloud.
The journey began with this idea of agile and how we become more agile and the realization that if we want to be on the forefront of technology in the financial industry, we need to be able to use the building blocks of technology, which means the separation of storage and compute and the ability to build api’s.
Capital One started with going to agile, moving to a developer environment that was continuous integration and continuous delivery and then also building micro services and api’s, so we have a good portion of the structure, that is now cloud native and we have focused now.
I am on the data on the backend and transforming that to the cloud which has brought us to snow, it’s been transformational for your company, it helps you be a technology coming off this company.
I think that’s one of the things that we’ve talked about before one of the things that drew me to Capital One which is that whole focus on being a technology company and realizing that Capital One is serious about this transformation.
So the engineer is working here, the greatest satisfaction for an engineer is having someone use your product right, so being funded and being able to deliver those products to our company, but to our customers it is exciting.
One is data-driven, we are very careful about giving our customers great experiences and I think the pivot for us, you talk about the journey in terms of our thinking, the customer is around this idea that instead of going to the bank to do your banking, we want to bring the bank to the customer.
The customer feels like their control, they know what’s going on and how I manage their money, but it’s them who’s doing it, so how do we make those experiences close to the customer moving beyond being able to do a deposit on your phone to understand everything about that customer and how we can provide experiences that make their life better.
So one example for us is the second look which is a product that is built on our data ecosystem and what this does is that it keeps track of a customer’s patterns, so let’s suppose Bob that I’ll assume you’re good tipper.
So let’s suppose you normally tip 25% when you go to a restaurant and your bills come in and your receipts and things and then one time you do a 50% tip, we don’t know if you happen to like that waiter or waitress or server.
If it was more about the fact that someone changed the numbers right, so we will send you that alert and then you can look at your phone and ask whether it is true and that is one of the products that we’re in the process of moving to snowflake now.
What have you learned from them coming to learning, the IT learning, I think probably the biggest learning is interesting that is a little bit orthogonal, but is network the change in networking and moving data over the network, that’s a big one.
If someone is getting started, how do they think about that? Because once you get in the cloud as you said, our data is on s3 the transformation in and out of these different systems is easy, once you’re up there, but that networking piece is like that.
Then I think for us it’s a whole transformation on that ETL layer and how users access our data and I think you think of how data has evolved and there are more different types of data available today, we are excited about being able to use these systems for machine learning.
That’s cool, so turn those machine learning algorithms on and seeing what we can learn, that gets back to your current environment, tell me how you use data today because Capital One is is an interesting environment, we talked a little bit about the first look, a second look.
There’s another interesting use case for us which is around fraud, so how can we use these big data systems? It’s real time speed and these queries that you give us in snowflake that allows us to detect fraud when it’s happening or even before that’s where machine learning comes in, we would love to be able to and we can in many instances alert you.
When you use a particular credit card in a particular place, we will also want to be able to do a fraud alert and allow you to make a choice, we’ve got in place now and we’re moving that to the cloud, but even be able to take that a step further and be able to detect fraud before it happens.
So you want to do that, what are the challenges that exist in your current data environment? What are the challenges there? So some of the big ones are speed to be able to access the data again, that’s a place where snowflake helps us out, because it allows us to work in much more close to real time.
Another big thing for us is the resiliency right, so as we are on Prem, you’ve got a whole different footprint, but think about when we’re in the cloud and that journey to be able to take us to that point where we can failover and I don’t want to failover when there’s a disaster, I want to failover on snowflake every month.
I think for us the other piece of that is that it’s very apropos for us now, we’ve had one hurricane, another one is coming down, but if you happen to be a consumer and you are in a situation where there is a need you want to be able to get to your money so that failover is so important for us to be able to make sure that people can get to their assets when they need them and get them quickly.
So one of these about your challenges that struck me and Linda from the first time I talked to the Capital One team was the sophistication of your end-users, you have a lot of end-users and they’re very sophisticated.
It surprised me as well, we have 9 over 9,000 sequel users in our company, I remember when I first got to Microsoft one of the training classes, I went to as a class on how to do sequel joins and there was this kind of pride around, I can do a sequel joint in here, we’ve got 9,000 business people from analysts to data scientists to others who can use data in complex ways.
We’ve been working together now and it’s early days on the VPS pod, we have the VPS pod up and running since early July, tell me about what your experience has been, it’s great, we look at the transformation and from multiple vectors, one of them is the users in terms of that.
They are engineers across the company, how do we train them up to snowflake and they’re anxious, they’re very excited, we have many of them coming on board, we’ve got our test environments up and they’re available for many people, we’ve got our card data loaded now.
We’re a credit card company in many ways, it is one of our services and we have that credit card data loaded up so that we can enable some of these scenarios that you and I have been talking about.
I think the other one is cool, as a company, we purchase good products like snowflake, there are always those integration points that you still need to develop internally and that gives the engineers an opportunity when you move to the cloud.
This is the difference definitely from on-prem in order to increase your size, you have to increase your hardware, so you’re very clear, you know what your costs are, but on the cloud, when you can scale out, your cost can be hidden, so one of the things we’re doing is that we’re launching.
At the end of the month, it comes a metering tool, that’s great, that will be able to be used by our finance folks and the line of business technical leaders so that they can understand their use and then we will extend that eventually to projections one of the attributes about snowflake, we do collect data about usage and we have all that data about queries and everything else.
So if there are any issues we can look back and see what happens, but it is all something that is accessible to you, you can do all kinds of things with that data, such as bill back, it’s a great use case, I think one of the things that I found interesting, it’s a very sophisticated organization with very sophisticated users.
You had a very solid infrastructure on premises, I think it was quite a strong infrastructure that was in place, what I heard from folks is that there was an insatiable desire for more and a lot more obviously, there’s a cost issue.
But we want to make sure that the business can be served, we see these transformations and industries, it’s because of new capabilities, this ubiquity of data everywhere gives us all kinds of opportunities that we didn’t have before separation of compute and storage.
We want to be able to leverage those with that, if we miss out on opportunities to go to this new place and we are an innovative technical company, what’s your vision? Where do you expect it? Where do you see it?
We see this going, then over the next few days, I think we continue to evolve and put that customer in the center of our part of the data experience and adding more data, becoming more complex.
Then I think the key thing on that question of where we are going is machine learning and our anticipation of where that will help us to understand the most important and innovative solutions, we can give to our customers.
I think it’s great to close this session with that about the customer, because it’s the end customer that always matters and data help us to serve our customers better and, as a vendor into this industry, I can say that that there’s nothing better than having the privilege of working with a team as high quality as the folks that we’ve worked with a Capital One.
It’s been an incredible experience and we couldn’t have gotten to where we are today, it’s snowflake with virtual private, snowflake without the help of the Capital One team, so I want to thank you and your team, you’re very welcome, we appreciate the partnership very much and thank you for your help today, thanks a lot, Linda.