Full transcript
0:00paler is a company that's valued at
0:02nearly a billion dollars but not that
0:05many people including those inside of
0:07Silicon Valley know what it is what is
0:08Talent here well in every large scale
0:12Enterprise you essentially have this
0:13problem that you have data in different
0:14databases and it's very hard for humans
0:17to actually interact with that data and
0:19what we basically do is we promote
0:21human- driven uh synergies between
0:23humans and computers by integrating
0:25every data store you have any kind of
0:26data and at any scale and we also
0:29provide privacy protection so that you
0:30only see the data you're allowed to see
0:32now this sounds like an incredibly
0:33boring problem but actually it's the
0:35core issue you have if you want to do
0:37things like prevent cyber attacks
0:39enforce civil liberties enforce
0:41standards on PRI of privacy on data or
0:44in some when some of our work actually
0:45sto Terror attacks so the core of it is
0:47to be able to solve all these problems
0:50through your paler platform right the
0:52DNA though of palent here is tied to
0:55PayPal one of your co-founders is Peter
0:57teal can you talk about the beginning of
0:59the company how you were able to grow
1:01this product from something that was
1:04created in PayPal so the key there are
1:07two key components to that one what the
1:09methodology that was developed at PayPal
1:11which was basically the use of human
1:13analysis to reduce fraud so they had
1:15this massive problem of of essentially
1:17cyber fraud and was putting them out of
1:19business they tried algorithma
1:21approaches so they you go and you get
1:22algorithms and you try to reduce the
1:24fraud by applying those algorithms to
1:25large data sets one of the interesting
1:27things about that is it doesn't work out
1:28very well because the the the opponent
1:31is highly adaptive you have an algorithm
1:33that finds this Behavior they figur it
1:34out they change what you need is a human
1:36mind that's adaptive against an Adaptive
1:38enemy or adaptive opponent so having the
1:41human mind apply its own version of
1:43algorithms to data uh and that actually
1:46was very very powerful in reducing fraud
1:49so since we knew that worked we marched
1:50off kind of naively to the intelligence
1:52community and said look we'll build this
1:53into a product now again to a lot of
1:56your questions the key moment here is we
1:58didn't want to do this as a service you
2:00could say we have a methodology we'll
2:01build every time we sit down we'll do
2:03this one we'll do it each time and we'll
2:04charge you we'll charge you about we had
2:06this idea that has run through our whole
2:08company that we will try to get this
2:09into a product meaning that we would
2:12solve the underlying issues that would
2:13work in eny Enterprise now what we found
2:15when we went to the intelligence
2:16Community is unlike PayPal uh they had
2:19lots of unstructured data the data
2:21stores were much larger uh they weren't
2:23built to communicate you had very very
2:25technical users and non-technical users
2:27and you had this massive issue of
2:29privacy protection ction so um and
2:31collaboration so in the PayPal context
2:33you allow any user to see all the data
2:36but in the governmental or even consumer
2:38context you can't allow end users to see
2:40every bit of data they only get to see
2:41the subset they're allowed to see so to
2:43take the PayPal model which would have
2:45been a oneoff approach that would
2:47require lots of Services hours and turn
2:49it into a product you had to productize
2:51the ability to integrate the data and
2:53productize ability but integrate the
2:55data meaning any kind of data not just
2:56simple structured data and that took us
2:583 years and and a very very strong
3:01engineering team it's a fairly
3:03complicated product and that's its
3:05strength but early on you must have had
3:07some doubts when you're creating this
3:09product would it work would it you know
3:11actually be used in the market when was
3:12your aha moment when you felt like okay
3:15we have something that's scalable that
3:17the government that other private
3:18institutions are really going to want to
3:20use well of course it was very scary
3:23since you know doing Enterprise software
3:252005 to 2009 was a little bit like you
3:29know starting a circus you know in the
3:31middle of pal Alta with Engineers it was
3:33not popular or it was popular with the
3:35wrong people mainly which was us and a
3:37couple investors like you know Peter who
3:39was a co-founder and so we didn't know
3:40it would actually work for till 2008 and
3:43we didn't know anyone would buy it to
3:45really mid 2008 so third quarter 2008
3:49and until then we were just operating on
3:51the faith that we had something really
3:53important now the real proof was we saw
3:56massive adoption without a Salesforce so
3:59this this is how we knew it was working
4:01because one person would email another
4:03in their they have classified networks
4:05and say this is awesome you have to get
4:07that and so one of the reasons to
4:09actually be very focused on an
4:10engineering team as opposed to a sales
4:12team is you really need to know is the
4:14Delta between what they have with us and
4:16what they could have really significant
4:19how difficult was it to break into DC to
4:22get the first government contract when
4:24you stepped into that first meeting what
4:26was it like and how did you make your
4:28case uh we did a very bad job making our
4:31case it was very difficult uh we didn't
4:35understand what they were saying they
4:36didn't understand what we were saying uh
4:39um I you know I think the first 100
4:42meetings or so were Frau with
4:44misunderstandings and uh you know we
4:46basically went in and said we have this
4:47tool we didn't understand their data
4:49sets we really didn't understand their
4:50problems we didn't understand their
4:52language they didn't understand ours we
4:55said from the beginning we're not hiring
4:57any people just cuz they're from
4:58government we're just for hiring
5:00Engineers uh many of the people in our
5:02company don't own suits still don't but
5:05the thing that resonated with them was
5:07we said we are not selling you a service
5:10we are not going to come and sell you
5:11Engineers we are going to sell you a
5:13finished product and we are going to
5:14show you it demonstrate its value
5:16against your data how many government
5:18contracts do you currently hold today
5:21well most of our contracts have massive
5:24Clauses in them saying we can't we can't
5:26disclose but I think uh the way you
5:28could think of it is uh if we uh we have
5:31280 people to really deal with the the
5:34footprint of where where we're at now
5:36it' be much better if we had 450 or 500
5:38MH so that gives you a sense of the
5:40scale do you think you're always going
5:41to stay true to that engineering is at
5:44the core of philosophy will you ever
5:45build out a sales or marketing team is
5:48that ever going to be part of the DNA
5:50Palance here uh I hope not I mean I
5:52could get hit by a car the core team
5:53could get by a car we could our
5:55investors could get tired of of the fact
5:57that we don't hire sales the thing is is
6:00if you are iterating on a problem that
6:01you want to be important 3 years from
6:03now it's better to have Engineers
6:06figuring out what the core issues are
6:08and then iterate against them if you
6:09want to optimize on Revenue next quarter
6:11or even in the next N9 months you want
6:13to be have heavy on Salesforce long
6:15sales short engineering we're long on
6:18the dealing with the most important
6:19problems that we can find uh dealing
6:22with them in a productized way so that
6:23they scale for the client um and because
6:26we're long on that and short on what
6:28happens in near term we are not planning
6:31to hire sales people we still haven't
6:32hired any we we don't really hire
6:35non-technical people very often uh and
6:37we don't have a marketing department and
6:39we're not planning to get any of them