Democratizing Robots with Daniel Dines (UiPath)

Okay Welcome Daniel thank you All right so uipath I think I read that Your company is now worth seven billion Dollars seven supposedly yeah supposedly Depending on how evaluate things but you Just you raised that round earlier this Year it was 586 million is that right yeah But more than half a billion more than Half a million Um so I think a lot of people here would Appreciate if you just told them how to Do that Well I I started early like in 2005 and then After 10 years of struggle we were able To raise a Tiny Seed round we from our Friends here in Germany early bird which Is a good fun and I learned a lot during My uh during my early You know days of raising Capital our First race took us 14 months to complete And that one probably wasn't uh uh a Half a billion dollars no it was like 1.6 million And then the I was less than I want to Call only less than two months to Complete from the first discussions to Money in the Bank wow okay so what tell Tell me a little bit about that struggle You mentioned you said they're 10 years Of strike like how was it in the Beginning because it's really been

Recent that the inflection point has Come on for your growth in terms of your Customer numbers and revenue so How do you what's the early years look Like and how do you keep going I guess During that time well before starting UI Buff I used to work for Microsoft in Seattle but uh I wasn't feeling so Comfortable there in the west coast and I had a crazy idea to start the company Right go back and I thought that going Back to Bucharest Romania would be the Best which is a stupid idea well yeah so What what about it though appeal to you The specifically what well I knew the Culture and I was able to hire some Engineers without paying crazy You know salaries but it was really Before knowing anything it was before Even TechCrunch like 2006 started to Watch it from Romania not from the US But I I made all the stupid mistakes Possible in building a startup like Failing too slow You know keeping on the product too much Right it's it's been really a disaster But somehow we build a technology on the Umbrella of computer vision And that we kept going and going with That stuff and we were really lucky to Find a very big Market yeah yeah and I Mean this Market let's talk a bit about That too like the the product uh is Robotic process doesn't automatic RPA is

The the fancy term Um but I guess can you just give us a Summary of that for people that might Not be as familiar with what that is Well I start a bit of how how we do work Today what defines our work it's uh it's Been shaped uh a lot during the 70s and The 80s you know personal computer and Then the graphical user interface and Then email productivity to a line of Business application so nowadays all of All we do is to you know read email take Some data from an email work in a Document send it to some someone else Ask from an approval but way use these Systems by the graphical user interface Right Business people make these processes if You have a manager and they ask you send Me every week a report about you know The stock exchange in Berlin right okay This is work yeah how do you do the work You do your research you use but you use The interface If you want to automate this work let's Say you will have to translate the work You do visually and to some apis but That's not an easy translation and it's Not a one-to-one translation and it Requires a lot of Development time product managers to Make to understand between the two Worlds the business analysts and the Developers

But our technology simply follow People's steps we work with the we touch The system via the user interface okay So I can train I can train our Technology to really replicate the steps That you do so you don't change the Infrastructure you don't have to Translate into a computer language you Use human readable interfaces it's the Equivalent of self-driving cars it's and It's basically use the same technology Which is called computer vision yeah so Is that then I mean you brought up Software in cars that's I think that's a Good kind of analogy uh for is there a Danger than that it's just a bridge Technology and that eventually you get To a point where perhaps if we have more Advanced AI like that there's you skip This step where you require this kind of Translation of the the human input into Machine input well we use first of all We use a lot of AI into implementing our Own technology computer vision is based On there are a lot of deep learning Models that we do behind the scene But uh I can tell you right now the it's Not only for legacy systems we use it Internally and we save to the tune at 100 000 hours per year which is the Equivalent of maybe 60 70 people at work Right our customers include Uber Google Facebook so it's it's a technology that Applies everywhere because it really

Taps into the nature of work yeah yeah I Was actually it's interesting you Mentioned that you use it internally so You dog food everything yeah and how Does that uh How does that work like is it is it Something where you find that your Culture readily kind of adopts to that Or do you find because I I feel like With something like this uh not that I'd Like doing these things but that I am Used to them and that it might be Difficult to adjust because people are Naturally static in their behavior right So so how does that work for yourself That's absolutely true and this is what We've seen internally you especially if You are in a company that grows a lot You started some with some a small group Of people they build some manual Processes they don't feel any any need To automate them right the frequency of Those processes is small yeah and Everybody just knows everybody just Knows when you suddenly have to grow and Then people that had multiple heads you Know separate one hit only to this Process and they hire a bunch of people And we find ourselves with 300 people in Our corporate group right which is a lot Only in finance to do a lot of I don't Know processing invoice processing and All this stuff oh yeah you get whole Departments that are just process

Managers or like sub departments that Are process managers and we didn't and Really we didn't really think of Consuming our software because we had to Scale so fast because it's faster to add People to the problem but that was wrong Inherently wrong way and now we we Started for you know like past 18 months To really push brutally our technology Within our company and I told everybody This is not to replace people this is to Make your job better but so on that Point like you did you did have layoffs And it sounds like the layoffs are Coincident with when you've realized Some of these you know efficiencies Right is is that directly correlated Like aren't people actually losing their Jobs or let's get into why the playoffs Let's say I would I would go later in The in the nature of Automation and its Relation to job losses but I can tell You about us where we scale from 2015 From 10 employees Into today to 3 000 employees way higher 2 000 people in the past 18 months right Beginning of this year Roy worth 2 000 People we will end the year with 3 000 People it's kind of natural when you Scale as fast to have a lot of issue First of all in the organizational Design a lot of overlapped jobs right That you know different people You end up creating silos and at some

Point you have to be able to change the Organization And uh we had a lot of we I being you Know being able to do an IPO was really A big initiative but I discovered that Actually some of our own people used the IPO as a reason to slow down our Processes so we stopped becoming really A customer first company So now that was the one of the primary Reason of this reward get rid of you Know bureaucracy streamline processes Becoming much more agile and I'm really Happy with What's going on in this run and second You need to despite the thing that we we Uh we have hundreds of millions in the Bank we have to be in control of our Destiny so we want to be we want to be Profitable you know in the short Horizon Of time right in 2020 our aim is to Break even So so one of the aims you mentioned There was was toward in service you felt Like it was a response to the plan to go IPO because you had planned what a year To two years as of the beginning it was One of the reason yeah we have Introduced bureaucracy and you know a Lot of redundant jobs yeah so is that at All related to your cfo's departure Because you hired the CEO CFO uh Marie Myers I believe at the beginning of the Year and then you know she's she's now

Finishing out her term it was that an Incompatibility that you're talking About related to culture or is that well I I wouldn't say it's an uh it's related To IPO Readiness or specific things okay But what I can tell in a in a hyper Growth Company It's normal to for executives to depart Fifty percent executive chair it's Normally in a company I really like Marie and we had a very friendly uh and Positive departure and we but That was not she was not the right Choice for uipath that was all she's a Fantastic great individual but it Happens yep it's many companies it Happens yeah so what about the IPO plans Are those still because you mentioned Taking control here on Destiny is that Still the ultimate goal or do you Anticipate going back to the market for For more funding privately prior to that Or what well we we might it uh it really Depends on the market conditions around The IPO we plan for an IPO of somewhere In 2021 We'll uh we'll really see about it Market is still very strong the private Capital Market can offer really high Valuation that is right now it's they Are very close with public valuation so IPO is just another fundraising event And it's a good milestone for for us and For for all our employees

So on the fundraising too I I heard a Story or read a story that you refused a Billion dollar in funding from uh South Bank from South bank's Venture Army yeah Is that true can you well I I met Masa I Mean it was funny he's very he's an Intense man oh yeah yeah and I I went to I went to him and I was a bit right I Said we don't burn we really don't burn Too much money we have a very efficient Business nothing he said to me this is Stupid you should burn a lot of money Take a billion now from me burn burn it And come next year and I'll get I'll Give you another opinion it doesn't Sound so wise in retrospect well at that Point that was August 2018 so at that Point still it was uh it looked wise but Yeah everything was Rosy then right Uh so let's talk a bit about uh kind of Future plans product wise yeah so we Talked about future Financial of the Company but like what what are you what Are you looking at next what are you Going to tackle next we talked about the Acquisitions are you going to make more Acquisitions or yeah I'm more I'm really Happy with the acquisition that we made Because it allowed us to extend from Pure RPA well end-to-end automation Platform so we plan we we have tools That help our customers into the process Discovery phase which is kind of a Bottleneck so we we have acquired a

Process mining company that we are Integrating completely into RPA Dutch Company we build our own Stuff because we use our own thank you We use our own computer vision To Discover What people actually do right there and Do some good recommendations for Automations We uh we also have launched a tool for Business analysts Automation require at least our existing Tool require a bit of programming [Music] Like some kind of expertise or knowledge Yeah not huge but some but now we've Built a tool for business analysts that Requires just to be proficient in a tool Like Microsoft Excel yeah they will be Able to automate simple activities yeah And we've seen really fantastic adoption Of this of this new platform and is that Deployed that's already shipping that Yes yes it's in uh it's in public Preview and you plan to kind of expand How sophisticated the tasks are that can Handle yeah yeah exactly exactly right Now we play that it's the it's bringing A lot of cognitive Technologies to be Able to extend because naturally our Technology being emulating people we Have to progress more into cognitive Things right right because you have to

You you if you look at all the invoices In the world once you once you've seen a Few invoices yeah I can show you Whatever invoice in English and you'll Be able to tell me what are the lifetime Everything yeah you don't have to I Don't have to train you to give you Templates and anything this is how we Learn so we build machine learning Models to become smarter and smarter to Work like people right I this kind of Ties back into one thing we step away From but what about you know criticism That it is a job replacement thing or That because you're doing this instead Of having so coming back to this yeah Automation very rarely was able to Replace jumps because people don't People rarely do simpler tasks right They their job Is compounded by a series of tasks right We call them activities we are able to Automate some of these activities but It's very rare that you can automate a Job in itself and I've seen a McKinsey Prediction that only five percent of Jobs in the world now can be fully Automated by the state-of-the-art Technology But technology can change jobs this is Very important and can change jobs for Better I'll give you a totally unrelated Example from farming and we've seen a Startup here in the look think how

Farming look 100 years ago it was a Really tough job very mad you know Farming is fun now you look only with Screens and you have you know GPS like a Video game it's like a video game it's Fun I like to be a farmer no it's fun to Be a farmer we we will make jobs to be Fun again Oh great okay so you're talking about Like essentially changing the quality Right the quality of how time is spent Yeah we're we're seeing we take the Robot out of human because nobody wants To do the activities that we can Automate right and how do you help the Uh we we touched on this a little bit Too but how do you help the uh Transition of the worker right who is Who is used to doing this thing and who Also still has to do that task to some Extent to contribute to the learning of How to automate it right but how do you Then uh reorient them so that they know How to take advantage of the new time They have or you know retraining I Suppose well first of all many this uh There there will always be exceptions That only humans can deal with so some Of some of them instead of spending time Doing the repetitive part right we'll Deal only with the exceptions maybe most Of the invoice may be paid automatically But some of them require calling the the Customer someone it's much better to

Spend your time calling the customer Offering a good experience yeah yeah and This is our this is our experience People are moving from more of a really Process oriented jobs into more customer Focused jobs okay yes it's better for It's better for their customers better For employees better for everybody Okay great so what uh what do you see in Like 10 years time what is the work Force look like in terms of automation How much is it penetrated I guess Well to me 10 years is really a big time Horizon I I never thought more than six tool two Years so okay let's give it two years Out three years I think uh I think that In uh in two years The big top 10 000 companies in the World will will will seriously adopt Automation as a priority and the shift In this wave cannot be stopped and People will uh people will understand That robots are their friends it's a lot Of Terminator kind of culture in the Western world that robots have your Enemies actually robots are your friends You have a robot in your kitchen you Have a robot that is you know cleaning The floors yes yeah yeah okay well That's pretty Rosy I'm looking forward To that uh so thanks I think we're just About out of time but I appreciate the Conversation that's great thank you

Thanks guys Thanks