Mastercard - Using Alteryx to Drive T&E Compliance: Priceless - Inspire 2017

In a complex and global company like Mastercard, ensuring the accuracy of company travel and entertainment expenses is critical. Thanks to Alteryx, Mastercard's Shared Services organization has been able to improve spending visibility, identify spending patterns, and improve cardholder behavior. In this session, you'll learn how Mastercard utilizes Alteryx to drive T&E compliance and identify patterns in spending while also utilizing forecasting models to determine the estimated cost savings and overall impact of travel and expenses on the company's bottom line.

Video Transcription

Rhianna Thomas:
Thank you! Hey Everybody! So I'm Rihanna Thomas. I was a senior data analyst, data scientist for MasterCard and their shared services. I actually recently accepted a new role in a new company. I've been there a couple of weeks now but I'm still here presenting for MasterCard on the project we did.

No, I can't take care of your MasterCard debt. I'm very, very, very sorry that's out of my capabilities now. Natalia was a co-worker of mine, she sat on the same team and so she is to present about a couple of things to talk about since I'm no longer with MasterCard and then I'll take over with the project.

We're going to through MasterCard in numbers. Just give you a little brief history of MasterCard. How many people we employ, you know, how many countries we're involved in. Things like that. Why we chose Alteryx, and then our challenge we had with T&E, which is travel and entertainment.

How many of you here have a corporate card? Majority of us do right? Yeah, exactly. So you're going to learn a little bit about MasterCard and the fact that, you know, we have quite a few employees.  Which means we have quite a few corporate cards we have issued and then we're going to talk about the steps we took. We call them steps one through 100. I'm not going to go through every single step. Then, how we visualize. We do use Tableau. We have an enterprise version instance of Tableau set up. We have a server and things like that, we'll go through that a little bit. Then we'll talk about forecasting, could we forecast our impact? Could we figure out, you know, what impact we are actually going to have on the company? How far, you know, out could we really forecast what our impact will be? Then we'll leave some time at the end for some Q and A.

A little bit about us, I love to travel! I tell my coworkers I work because I like to travel and you have to have money to do that. I'd like to take one big trip a year. Last year my big trip was to Australia. That's just a little bit about me.

Natalia Terhard:
Thank you Rhianna! My name is Natalia Terhard and I also like to travel and I like scuba diving! Very often I combine these two hobbies since there are not so many places to go scuba diving in Missouri where I live, and there is definitely no ice diving. This picture was from one of my trips in Russia.

Now we talk about MasterCard. I'm sure many of you have heard about the company? Yes? No? Yeah, a few of you. MasterCard is a technology company. We do not issue cards but we develop advanced payment solutions and seamlessly process bills of transactions around the world every year. We connect consumers, financial institutions, merchants, governments, worldwide and we enable them to use electronic forms of payment instead of cash and checks.

We use technology and data driven insights to make electronic payments more secure, more convenient and more efficient for people everywhere. Our business has a global reach, extending to more than 200 countries and territories and it continues to grow in the world where 85% of transactions are still made in cash and checks.

Our analytics and automation team is part of the shared services center at MasterCard, and we support transactions, accounting and billing services.  Like many companies we have a lot of challenges. A lot of excel reports, analytics styles, high dependency on a small groups of analysts. Data blending was difficult and time consuming, and we spent about 85% of our time on data preparation and only 25%, excuse me, 75% on data preparation and 25% on analysis. We needed to fix that and move away from reporting and start analyzing data. We also wanted our business user, subject matter expert, to drop excel improve the efficiency in preparing data, and also start analyzing data.

We're very happy that some of our colleagues joined us for the conference to find their inspiration with Alteryx. We need to find a common platform that we would all work together on and Alteryx was the perfect solution. It's a user friendly platform and doesn't require coding experience. We can schedule reusable and repeatable workflows and never touch them again. And it also perfectly integrates with tableau, which is very important to us because we build a lot of dead boards in tableau. Just to give you an idea of our implementation journey, it was not quick and I believe many of you can relate to that.

We started with one license in September 2015 and we built a case for our enterprise architectural team who guided the company in IT infrastructure.  In a couple of months later we started proof of concept and bought four more licenses. Then the next year we made really good progress automating reports, building our workflows to build and refresh data sources. Most important we made good progress with the T & E case we'll be talking about today. One year later, we met again with enterprise architecture and we proved that no other tool had all the capabilities of Alteryx platform. Finally, we bought Alteryx server in December last year. Currently there are fifteen Alteryx users at MasterCard and we are about to buy fourteen more licenses.

We have built about 80 workflows and this is just the beginning. It's a very exciting time for us because we know the opportunities are endless. With that I will pass it over back to Rihanna.

Rhianna Thomas:

Okay so we had a problem. It wasn't really a problem. It's just something that we never looked at. We have, I say we, MasterCard has 11,000 employees. Out of those 11,000 employees 8,000 of them have a corporate card. We have a huge, huge, huge corporate card program. With that we do, we have presence in all over the world right? We travel. Most employees travel at least once per year. We have lot of sales people. They're traveling constantly. Traveling about the world constantly. With that we have about a million expense lines a year. We swipe our corporate MasterCard a million times a year. 

[00:06:57] No one can look at a million individual records right? We all know that, that's completely impossible, we can't do it. You cannot have a team of people who does it. With that, we had this thing where we have all these transactions. We're spending all this money per year in trave and entertainment, no ones looking at it. No ones looking to see, I mean yeah we fill out our expense reports and things like that. You know you provide justification, your manager approves it, Half the time, let's be honest, you're the manager you're like yeah approved, right? You looked at it for all of like 30 seconds.

Took longer to log into the expense reporting tool than it did for you to approve the expenses. We needed a way to look at these expenses to figure out, are they legit, right? I heard people are getting hotel rooms for their family members when they come to town. Are they taking their friends out for cocktails. Are they legitimately using their corporate card for business expenses.

With this, we also decided we wanted to look and see how many of these transactions followed actual corporate card policy. How many of these transactions were potentially misuse of this corporate card. With the card policy every country kind of has its own flavor. We have different rules. Different regulations that we have to follow for that individual country. It was really being able to look at a million expense lines per year and really trying to be able to hone in country by country and by region determine are we using the expenses appropriately.

Where do we even start, right? Step one was we were gonna leverage Alteryx. We had this great idea that we would use Alteryx to comb through all of these expense lines base dona set of rules that we kind of came up with. The first step was we needed to connect to Oracle. All of our expenses flow through oracle. We have an also work day, which is our HR reporting system. Employees names, you know, where you're located, where you work, your manager, that type of situation.

We need to be able to blend these two things together. We need to be able to have a comprehensive picture of the expenses, the employee. Then we thought, why not go ahead and get the data information too we have from our travel provider, which is Carlson Wagonlit Travel. Anytime a MasterCard employee is going to travel anywhere, we have a travel agent partner. We go in, we book our flights, our hotels, that type of situation. Well, in Oracle, our expense lines, we never know the original intent of the employee. We don't know that you went in a booked your ticket five days in advance, or 55 days in advance.

We do have policy around it, we have rules we are supposed to be following and things like that. We thought if we could blend our Oracle data and our HR data and our travel partner data, and get one large comprehensive overview. Everything from the employee booking the plane, train, and automobile ticket to where they ate breakfast, lunch, and dinner while they were there. Then that would actually give us a comprehensive picture of their entire trip. Potentially the intent of the trip. How many days they were there and things like that.

Now we have million records right, and all this information. What do we do now? We need to manage it, we need to make it something sizable. We started with a basic rule set. We had four rules five rules, four rules. We thought, lets just run all the expenses over the last year through these four rules. We came up with a key word search. Our key word search in the beginning was basic, we were looking for like cake. How many people buy birthday cakes with their corporate cards, you know, for their coworkers. How many people are buying visa information, not visa the credit card that's like a no no word. If you want to travel India or things like that we do have to have Visa and you know things like that.

[00:11:09] Then starting to looking at submission reports approval time-line. You know we have a policy, you have to have your expenses submitted by X number of days. Your manager has to approve them by x number of days. Stating to look at were employees following this rule. What was the time-frame between the submission and approval. You know, was it on the very last day that the expense was supposed to be submitted and it got approved or submitted then approved within a matter of minutes, things like that.

We started looking at weekend expenses. For the most part as a company, as employees, we don't travel on the weekends. We travel Monday through Friday. There are exceptions right? Especially if were flying internationally or things like that. We might fly in couple of days before, you know, fly home a couple days after and things like that. For the most part we don't really travel on the weekends.

Then we started looking at final expenses, where their final digits were zero zero, 100, 200, you know, were people rounding things. Were they maybe rounding up something. That's what we started to do. Once we ran all that through we got an initial output. All of these expenses were flagged for these four reasons. Some maybe were flagged for more than one reason and that's kind of what we came with.

Then after that we came up with this. This is our final Alteryx flow. My manager at the time joked if I got paid per tool I would be really really wealthy right? They don't pay me per tool I use so that's its kind of a bummer. This is where it starts, we have an initial date it was set over on the left, which is our connection directly to Oracle. What it does is it then goes from there, we have a couple of rules, a couple of things we take out, a couple of people we exclude and things lie that. We don't look at their expenses. Then from there each container you see is an individual rule, potentially a standard deviation rule. Was the expense more than two points above the standard deviation for Starbucks. Was it maybe more than two points above the standard deviation for an expense category such as a restaurant, coffee, hotel, things like that. Are we, some excessively spending. Are they always taking a suite when their at the hotel. Are they always upgraded to first class when they fly, things like that.

Then from there we work all the way through to the end to where we actually apply scoring methodology. Each expense then gets spit out into an individual line for what it was flagged for. Then we have a different waiting system. Some rules have a higher wait than others, they're more important. In a sense that maybe they are violating actual card policy. You know, its actually written in the policy, don't do x, they did x, so it's worth a little bit more than the final zero type situation because we cannot really say it was an illegitimate expense.

From there, we have 27 rules. We have 27 final rules that we have. That every expense flows through. Like I said we do exclude a couple of people just because they may be at a certain level. Their expenses are always the outlier for their travel purposes and things like that. Every expense flows through. We do it for the last 100 days. When we go back we have a look back and it pulls all expense over the last 100 days. It runs them through the tool.

We were able to integrate the Alteryx Spatial package. This is probably one of the coolest things I got to do with project when I was building it, was integrating Spatial. I had gotten this idea that, well I had seen a demo, right. I saw demo of how Spatial worked and that we could calculate the distance between two addresses, two points, something like that. I thought that would be a really cool idea, right. We are MasterCard so we know where the money was spent right. We know the address of the merchant and then from there we know where the employee works. We know their home zip code and things like that. Could we start looking at and saying well why would you spend money a mile from your house at Target, that doesn't make a lot of sense. Why would you have gone out to dinner for a work dinner that was a mile from work, from the office, it doesn't make a whole lot of sense.


We were able to insert that into the Spatial package, got that information, plot two points and then start calculating the distance between where the employee worked and where their home address was and where their corporate card was swiped. Then it kind of bright about some other questions of course, as everything does. From there we kind of started honing in on exactly what was too close to home or too close to home, things like that. Of course again there are always caveats to it. There are always people who fall outside of that spectrum. It was very cool, very cool rule we added in.

We also integrated the R tool. We had a summer intern, last year. She was from UC Berkley and she loved to write in R. We were trying to teach her Alteryx and teach her that there are other things to use but she loved R. She wrote very cool code for us that we could insert keywords and loop through expense justification, expense report, you know naming conventions and things like that. To actually look through and find key words and key phrases. Then what we do is after all the expenses run through the module. What happens is their output into a database. We use and Alteryx database, nothing super fancy or anything. It is just output into Alteryx database but then that allows us from there to move into our next phase, which is actually reviewing the expense line. From there we really wanted to start reviewing and looking and determining which rules were actually corporate card violations, like policy violations, and which ones were actual potential misuse of the card.

Then we review and we review and we review some more and we review some more right. We have like over a million expense line. Even for 100 days we could have an output of 15 to 20,000 expense lines that actually have to be individually reviewed. Our initial module when it ran, it flagged 4,00 expense lines for further review. This is, they were flagged for a reason. They hit one of the rules and so they are dumped in the database. From there we are like okay what do we do with it all now. This is cool we have all this data, what do we do. We decided that we needed a tool that would allow our T&E team and our corporate card team to actually look at the individual expense. Pull them up in Oracle. See what the justification was, maybe look at the overall tone of the expense report and things like that that. To determine was there follow up that needed. Do we need to contact the employee, maybe ask for some further explanation. Maybe inform them or educate them on the proper way to use the corporate care.

We cam up with an access database and a share point list which we thought was the best solution at the time, so that is the path we went down. It worked, it worked for six months. It worked for a whole six months. Then we started noticing that we have limitations. Our share point list are only allowed to be only x number of lines. The access database was starting to kind of chug a little. Especially for our team who reviews its expense to actually sit in South Africa, it was just slow, it just was not a feasible solution. We said okay we'll try it again, like lets figure our a better solution.

We went back to the drawing board. We needed a tool to review expense lines, to add notes to, things like that. At the time I had kind of been asking and begging for an Alteryx server. You can only beg so much, and they stop listening to you. You gotta beg then back off a little bit. Then you gotta beg again then you can back off for a little bit. I finally said, look guys this is an excellent reason we need the Alteryx server. This is broken and it's not working right. If we had the Alteryx server we could do it. They said all right, fine, go ahead and buy a server. I was like sweet! All right! That opened up the door for tons of other things right. Of course now we have like 25 workflows scheduled on the server to run and things like that which is awesome.

We decided, we being me, that I would build an application. This is gonna be great right. This gonna be easy peasy and it will take me like an hour to build an application. Well I was surely mistaken and eight days later, and not leaving this one room at work. I finally had an application built. What I wanted to do was I wanted the reviewers to actually be able to log in the server, to actually review the expense lines and everything just be houses and set on the Alteryx server. They never had to touch or open up another program and that was kind of the goal.

We kind of started using Alteryx server in maybe a more innovative way because I don't know that is was completely designed to do this. Of course we started to realize some limitations with it but what happens is we have a user interface. This is their teeny review folder here. What they can do is they can actually click it. They click the application and it opens up. It go ahead and grabs the first expense. Their set of expense for one employee. From there they get a questionnaire that pops up.

They can put in the expense line ID. They have their initial analysis. Was it compliant, or was it potential violation. We don't want to say it's a violation yet or non-compliant yet. They can put in their review notes. They can say what the expense was flagged for and then who it was reviewed by. Our team of reviewers is at the bottom, at the very bottom because I test a whole bunch and then I was like oh no I don't know which expenses are reviewed because I make you random things when I'm doing it. That's kind of how that works.

The review is complete. They go through, they review everything individually, line by line and its given a status. The expense was legitimate and there is no violation. They found reasonable justification for it. They found that receipts were attached, things like that. No further information. The expense that I highlighted are our corporate card policy, so the expense gets a status of a violation. The employer receives a confirmation email that just kind of says, in more business terms, by the way hers the policy. We state it for them, kind of like people don't do it again, you know what I mean.

Kind of like a little warning. We also have a further information is required, before the stats can be determined. We actually will email the employee and we'll say there's this expense can you please provide further justification, further information. Kind of doesn't state anything specific because we have found that they will be open and honest about it if we don't specify exactly what we are asking them for.


From there what the employees and reviews can do is once that email is sent out, the employee responds and they have another module they can go in and do the case review. The final expense review. This power point was created before were finished with all of the modules and they were finished right before I left MasterCard. Now what happened is they have some further modules to now where the employees email is automatically sent. After all of the expense lines are reviewed, what happens is they can run, they just run another application. What it does is just summarizes all the expense that reviewed for that one employee. We don't want them to receive like 100 emails on expenses we want justifications on. Then from there it actually just pops out the email to them automatically using Alteryx. That way it save our team a ton of time.

Our final step was to visualize right. Now that we have this process and we have this project completed in terms of how we are going to review everything and look at all the expense, of course our controllers and things like that. They were like well I want to know, I want to know what this country does. I want to know what this team does. They wanted very specific. We have an Alteryx and Tableau which gives us our nice little magic. We have multiple different dashboards, for multiple different audiences. This dashboard here that we're showing is an example of our operational dashboard. Our had of T&E needed a way to know how many expense were being viewed per day. Which rules, which flags were they were reviewing. Were they picking easier ones to review, were they skipping over anything. Then we also needed to create a controller and country mangers dashboard. They needed their own operate set because they needed to look at the expense from a different expense.

Then we have our executive management dashboard which is for our CFO. She wanted to see things in an overarching perspective from a company wide point of view. She wasn't as concerned with individual pockets.

Then we also wanted to quantify. Impact analysis is education required. Are certain areas prone to certain behavior and who is gonna benefit from seeing this. I think the biggest thing we kind of took away from this is education is required. A lot of people didn't know we had a corporate card policy. They just gt the card and that was it they were done. We set up a training. We set up like a walk me through type training. It was pushed out to all cardholders.

Our final impact, I'm gonna say this is October to February. October 1 to February 2017. We had 31,00 cases, we have 14,00 expenses that were reviewed, and we had 3,100 employees that we impacted. Considering we have 8,000 card holders were at you know 40 to 45% of our cardholders were impacted in this in one way or another.

Could we forecast? Could we actually forecast the impact to see if all this time and energy and money worth it. Can we determine how much money we are gonna save from actual misuse of the card and company funds. Could we actually see a change in behavior spot patterns. Did on person on one team get an email, so maybe the entire team started to behave a little better, you know clean up their act more type situation. Do certain groups or regions stand out? Where will this look like in one or two years. One or two years portion I cannot answer anymore because I am not there. So good luck to my teammates.

What was gonna happen. Could I forecast what was gonna happen. We started looking into time series forecasting in Alteryx the ETS versus Arima models. What happened is we got in, built a very nice forecasting model and we really don't have enough data yet. We only had at that time three to four months of review and it just really wasn't giving us the best output or solution.

We will begin to analyze the impacts by region, business unit. Things like that. Then external factors that affect spend that we cannot account for. I think a lot of companies are don't he push to of course spend less, spend less, so we cannot say that the overall reduction in these types of violations are non-compliant is really just because of us. It could just be that potentially they are spending less. Something we are going to have to take into consideration, spend over 2014 versus 2016 and then 2017. Then starting to actually look at the impact the difference there.

What's our future, what is future of T&E compliance at MasterCard. They will continue to develop new rules. There is a list probably of 100 rules that people have come up with that thy would like to implement and things like that. Then they would like to begin customer rules. I mentioned before there are country specific laws that we have to pay attention to. India is a good example, we can't always use your corporate card in India. A lot of times you have to take cash and things like that. There are a very cash heavy society. We are starting to look at things on a country specific level.

The expand expense review. Expand it on the interface on the gallery. As I've mention there a couple of things we've run into using the Alteryx server like we have. You can only have one person reviewing expense at one time because don't want them to overwrite each other's work in the database. Looking into implemented a SQL database to where we can actually lock records and things like that will be the next step that they take.

They want to start expanding the internal use of the results. Partnering with the internal audit team. Their initiatives on both fronts to review this type of review and analysis. If one team is doing it, we don't need to duplicate efforts. That would be a joint effort to work together. Then expand the forecasting models, really starting to hone in and see if we can actually forecast the impact and the difference we are going to make.

That is the end and time for any questions.

Crowd question:
Thank you Rhianna and Natalia. We have a few questions

Crowd question:
Hi, you said that you integrated Spatial analysis. What type of Spatial or special analysis you did in this?

Rhianna Thomas:
Sure, so we used the cast address system to get our points for the maps and then we calculate the distance between the two. We calculate the point for the merchant which we have the full address, the street address. Then we usually use, we have the full street address for the employees office, work location, not their home. Their home we just do it based off their zip code. The work location we have the full street address and then we calculate the distance between two points, and we say anything less than, I want to say it's like four of five miles something like that, we do thorough analysis on. A little bit further analysis on. We also set a dollar threshold in there too. If it's like less than 5 miles and over 500$ or something like that.

Crowd question:
You mentioned fraud detection early on, and periodically through the presentation but I may have just missed it. I mean we're in Las Vegas right. If I didn't miss it can you articulate some of the things that you did to implement fraud detection monitoring

Rhianna Thomas:
Yeah, the point they're are not is actually mining the data and looking for patterned. Spotting patterns and behaviors and things like that. The initial rule set and things we started with came from forensic accounting, from [inaudible 00:31:53]. They had a very brief PDF online about it. There is not a lot of information about it online. I don't think companies overly share what they do in this space. I don't know that we have today been able to detect fraud with it. We are always looking at this after the expense has already happened. It's more looking at are they violating policy or are they getting a hotel room that's two miles from their house because they are fighting with their spouse and they put it on their card type situation. No travel to justify it but there was no reason for that to happen. I think that's where we are at in looking at fraud.

Crowd question:
So you mentioned that you had 27 rules in your workflow and you had a backlog of maybe 100. How dynamic did you make your workflow? Say you want to add a new rule, do you have any tips or tricks implementing that.

Rhianna Thomas:
Yeah, sure so I will go back to the flow itself. Hold on. The flow is super dynamic. What kind of happens is you'll see there's like a box right here right. Everything on this side of the flow, listed everything from a time. All time I think we go back to 2014. What it does is we calculate the standard deviation of all time for Starbucks or for Cheesecake Factory or whatever. Then for the expense category, meal, meal self, meals business, things like that. Then everything from this part of the workflow over is only calculating things for the last 100 days. The reason we did that is, it was taking forever and a day to run this looking at all three millions expenses, running it through especially the R integrated package and the Spatial package. It was just taking way too long.

All we have to do now is when we are ready to insert a rule. What I will usually do is build it in a different workflow in a container. Then you just, it just pops in one after the other. Then it just joins in unions back up. It is super simple to add a new rule in. Depending and if you want to look at it from beginning of time in 2015, or if you just want to look at it for the 100 day time frame.

Crowd question:
My questions is on the analytical app that you built. The expense review app. How did you approach security so that one manager could necessarily see another managers employees expenses.

Rhianna Thomas:
Sure. The application on Alteryx server, the only people that have access to it are the T&E team. They are the only ones that can actually see an expense being reviewed. On the Tableau server, on the dashboard, which I could think where you could be going with the question. We use real level security. Our Tableau server is based on our actually directory. We use real level security to create a model. It's essentially a table calculations. What it does is look at the user that's logged in, finds them into the HR tree and that's how the real level security works in Tableau. Its a calculated field though, if you need it I can send it to you.

Thank you, any other questions.

Crowd question:
For your workflows that you built in the gallery. An you talk a little more about, it looks like you have a couple different workflows, and kind of what that process looks like. I am guessing one is just to write into the data versus the others

Rhianna Thomas:
Yeah, there is one called refresh expense line review. What happens when that app runs. It actually goes back through. It actually runs the huge module that you saw. Then from here it writes the expenses that need to be reviews to a temp database. The once that is run and done we usually just tell the group reviewing the expense lines, whenever they are ready to do it just ste ad run it over night and leave it be. It does take maybe three to four hours depending on how bogged down our server is at that time.

Then from there we have the review expenses and employee review selection. The employee review selects, and again this is where we had to kind of get creative. I needed a way to have expense essentially pop up, so they could see the expense number, the date, the merchant, the amount, the justification, and an other pertinent information that we thought for that right. However in an app a pop up like that or a PDF in a chain app has to be that last app in the chain. It can't run another one after it.

What happens is they go in and they click the employee section. It goes ahead and grabs the first employee and all of their expenses that need to be reviewed and does a PDF popup. In chrome it opens another window, and then they actually go into the review expense and that's what this is over here.

Now, there's two for each and their named with their names because after they review the expense it gets written to a database that has their review notes and their initial analysis and things like that. If they are both in there and they happen to hit the button at the same millisecond they would write over each others work. There are two ways to combat it. I was a fan of buy another Alteryx server because I could set it to just run one flow at a time. That would essentially put them in queue with one another. We could have changed our Alteryx server, we had to just run one flow at a time, but then that limits all of the other flows have on the server and we didn't want that to happen. The their solution which I am not saying is the best solutions by any means but it worked for the time being, was just to create two. This is going to become an issue if we have more than two people reviewing lines at a time. We hadn't hit that problem yet. As we become more mature with our Alteryx server, and things like that, I think it's a potential that we will buy another server and combat the problem.

Crowd question:
Thank you are there any other questions?

Crowd question:
You mentioned R, what was it actually used for?

Rhianna Thomas:
R? Yeah so its used to loop through a list of keywords. Cake, boyfriend, girlfriend, that could potentially be used in the expense justification. It loops through all of the justifications and key.

Crowd question:
Was there a reason to use that instead of using regular expressions?

Rhianna Thomas:
Our intern loved R, and I wasn't gonna argue with her. She was doing me a solid by writing that portion of it so I was like go for it. Secondly when we do that, we just keep the list in a CSV so if we add to it, its no like me actually having to go into the workflow to add that expression.

Any other questions? Okay with that I would like to thank Rhianna and Natialia for a great presentation and before our audience leave if you could pull up your app and rate this section that would be must appreciated. Thank you.


Experience the
Power of Alteryx
For Yourself.

Get Started