True AI and the Future of Networking: Rami Rahim, CEO Juniper Networks

0:03 welcome to another episode of the reboot chronicles a no holds barred form with global leaders authors entrepreneurs and

0:08 ceos about how organizations stay focused on growth and innovation in unprecedented times

0:16 i'm your host dean devious coming to you live from raviv's north american headquarters in chicago and we would

0:21 like to thank you for joining us from around the globe today [Music] i'd like to welcome Rami Rahim the ceo

0:27 of juniper networks to the reboot chronicles today headquartered in silicon valley with over 9 000 employees

0:33 in 50 countries and over 4.7 billion in revenue juniper and romney quite frankly

0:39 were pioneers in building the internet along the way he has um received a lot

0:44 of accolades my favorite one is he was named one of economic times most loved ceos i can't say that about myself

0:51 so uh that's a nice title by the way good to see you remy thank you dean it's pleasure to be here

0:56 i appreciate the invitation yeah we're big fans of um you know kind of the behind the scenes companies that

1:02 don't get enough attention um a lot of competition in your space of course but not enough talk about it i think so uh

1:09 you know juniper was a pivotal player in the creation of actually making the internet work something we all take for

1:15 granted that was back in the 90s you know from the dial-up era all the way through the broadband and beyond um

1:21 gosh at the time i was making the first cable modems so dating myself here but i think you were just a young engineer at

1:28 that time but let's just jump right into it um you know what from what i see my vantage point it

1:33 seems like telecom and infrastructure growth has kind of slowed i don't know about 22 23 with supply issues and stuff

1:41 maybe there's a lot of ordering going on but um i guess the first question right out of the box is how are you rebooting

1:47 into a more agile growth company and in how you position yourself against these competitors and kind of move from

1:54 this hardware centric world into more software ai driven cloud services

1:59 security 5g i could write a lot more buzzwords but just yeah how's that whole reboot how

2:05 you're approaching it and what's going to happen well what a great question to kick off

2:10 the conversation yeah i think jumping right into it juniper has always been a true innovator

2:17 in ip networking technology we were builders of uh the infrastructure that

2:25 helped the internet scale at a time when service providers were really struggling to keep up because it was really at the

2:31 cusp of the dot-com boom so imagine a little upstart company trying to convince service providers to trust

2:37 their networks with startups technology the only reason that i think they

2:42 managed or we managed to get that kind of trust is because they really had no choice traffic was exploding and they needed

2:49 technology that could keep up with that traffic and we had an innovative new architecture to developing routers core

2:56 routers at the time with silicon technology and a very scalable internet grade operating system

3:03 that could keep up with that demand and today you know we're proud to have

3:09 the top 50 service providers around the world as our customers but we have to your point absolutely by necessity

3:16 evolved our business and diversified our business so today if you look at our business it's actually very nicely split

3:23 between telecom operators which remain extremely important to us cloud providers all of the major

3:30 hyperscale cloud providers in the world are our customers and global 1000 type

3:35 enterprises that are increasingly trusting their networks on our technology and

3:42 i think in the early days our claim to fame our credibility our pedigree was in scale and performance

3:49 and of course that remains important today but i truly believe that the game has changed i think you know things have

3:55 evolved beyond just scale and performance into experience

4:01 right now if i talk to a cto a cio or ceo of a big fortune 500 company or a

4:08 telco or a cloud provider their biggest challenge is complexity in network operations that's slowing them down

4:15 that's in sometimes just bogging down progress and the solution to that is in

4:21 fact software driven innovation ai driven innovation that will remove the

4:27 complexity from running networks in fact to remove the human factor from running

4:32 networks altogether so if you look at our innovation today it is mostly in software

4:38 extra for example is our solution to day two automated operations of the data

4:44 center our missed ai engine is all about automated operations of client-to-cloud

4:50 enterprise services so juniper has evolved a lot from those early days of helping to scale the internet

4:56 you know i'll jump into that in a second the ai part but uh you you you as this offense happens on this program i have

5:03 flashbacks so at the time while you guys were coming out it was you and son and i was running the imagination network one

5:09 of the first online games companies and we had to build our own server rooms back then and you guys were you know we

5:15 weren't that big and you guys were supporting us and you know it was like startup working with startup and then there was sun who thought they were

5:20 going to rule the universe and it was it was an interesting time very hardware centric a lot of people a lot of we

5:26 called it sneakernet right a lot of people running around and and the industry is it seems like it has

5:31 evolved so much um so every time people hear ai it seems like

5:36 it's hyped up just about as much as metaverse um in terms of keywords it's it's it's

5:43 often misunderstood for instance from ravi the sponsor of this program you know we have an ai uh technology

5:49 that is at the edge in the at the edge for retailers and brands to help communicating personalized services with

5:55 all their customers whereas you have ai you know at the back on the back end and

6:00 in the actually all the way out to edge i'm sure but why is ai i guess how is it why is it so important sounds like

6:06 you're going to eliminate some some um manual processes and and and

6:11 and how is it actually going to help the industry you know at the network level well i'm glad you're asking because

6:17 honestly this is one of my favorite topics this might get too geeky little audience warning here

6:24 i will try to minimize the geekiness but you know being i was talking about myself

6:29 it's okay you know to talk about ai i think

6:35 you need to start by talking about data data is now the most precious

6:41 valuable resource in the world and we as a planet are accumulating data at an

6:47 unprecedented scale i heard a stat recently that 90 of all data in the

6:53 world has been accumulated in the last two years alone so now the question is what do you do

6:59 with all this data right and this is i think where ai comes in there is no way

7:04 for human beings to process this data manually ai is this new breed of learning that's

7:11 not learning through programming algorithms it's actually learning by looking at patterns in data

7:18 and now attaching it connecting it to our industry we are we as in an industry our customers

7:26 are sitting on gold mines of information that today are largely untapped

7:32 so ai as applied to our industry of networking is

7:38 incredibly promising it gives our customers the ability to collect data in real time i'm not talking about the

7:44 spooky customer you know payload stuff just performance data that gives us real

7:49 time visibility of the experience of every user on the network as they

7:56 do whatever they need to do on that network you take that data you learn from that data and then you automatically

8:02 translate it into useful things and for us those useful things those compelling

8:08 outcomes that we're able to offer our customers from this data are things like

8:14 the fastest deployment times the fewest trouble tickets the fastest time to

8:19 resolution of trouble when there is impact trouble typically you're solving problems before humans even know that a

8:26 problem exists um for those of you listening that means when your slack channel goes down this

8:32 can actually bring it back up quicker it identifies the problem and fixes it yeah and you know

8:39 you touched on something which is that there is a lot of ai washing out there there's a lot of

8:44 confusion there are a lot of people that are claiming ai when it's really just powerpoint ai

8:49 but i think we've moved from powerpoint into actual

8:55 real outcomes that we were able to offer our customers and probably the most fun

9:00 part of my job today as a ceo and also a technologist is to convert those

9:06 skeptics almost every day now into ai believers by demonstrating the

9:12 value that we can deliver through automated operations and i think you know honestly the future of networking

9:20 is about ai we are simply scratching the surface right now of what is possible

9:25 yeah and you're not just talking here you put your money where your mouth is you've made a couple acquisitions in the space we'll talk about how you dance

9:32 with startups in a little bit but you acquired maybe one or two ai companies to get ready for this this isn't something

9:38 you're just uh putting in your slide where as we say we absolutely have yes you know we've

9:43 always been a big innovative technology-led company we've all always invested a large part

9:50 of our revenue in r d but we're also you know humble enough to recognize that

9:56 sometimes the best innovation are outside of our four walls and we were very fortunate to combine with some

10:01 amazing team great talent and amazing technology mist is an example of an enterprise

10:09 architecture that is driven through cloud delivered ai and

10:14 actually the ai engine for mist is called marvis affectionately called

10:19 marvis marvis is honestly one of the most compelling technology breakthroughs

10:25 in our industry i have seen in my entire career and that's just because margaret's is

10:31 the first thing that i have seen that has really moved ai from a concept from you know just talk into actual

10:38 real-life results for our customers and we've taken it from there into appstra that

10:44 does something similar for the data center uh net rounds does something again similar for service provider wide

10:50 area networking i like marvis that uh that could trend as a new child's name we'll see if that

10:56 that picks up it's better than an acronym i mean come on so you've just hit on a bunch of things let's just try

11:02 to drill down and unpack a couple one is co-creation partnering with startups at kellogg i've

11:07 got a dancing with startups program we bring some of the largest and the smallest companies together on the planet really to help brands like you

11:14 figure out how to kind of dance that ecosystem who to partner with but it's always through a build by borrow lens so

11:20 you've talked about companies you bought which totally accelerated your ai capabilities and what about other things

11:26 in terms of partnering with startups on the whole co-creation side can you tell us what you're doing there

11:33 certainly i mean you know i i think if you look at juniper's pedigree our

11:38 history we've always we grew up as a challenger in an industry that had strong

11:44 incumbents and in order to succeed in that environment i think openness is your friend the

11:51 ability to demonstrate to your customers that you're not going to lock them into

11:56 anything you're going to embrace open interfaces um northbound southbound into

12:02 the optical air eastbound into peer solutions is very important to them and

12:07 that's something that i view as extremely important to maintain irrespective of how big we get i mean

12:14 many networks now we are the incumbents but i never want to lose fight or lose that

12:21 that that pedigree that we have in the company of openness because it's just so important for our customers now you know

12:28 it sometimes makes it more challenging for us because when you're this open you've got to constantly compete and

12:34 innovate and demonstrate to your customers that you have the very best solution um

12:40 but you know this is uh this is an area that i you know i absolutely always commit to my customers that having that

12:46 openness to work seamlessly with peer solutions with startup technologies that are out there is

12:53 a lasting uh element of working with juniper and you've got no shortage of

12:58 competitors as i mentioned and some of them you've kicked around the block i mean you know companies like cisco have struggled over the years i mean some

13:04 kind of kind of come and go depending on the ceos and the investments but it's got to be tough um because

13:10 some people call it the pizza box server market where you're just sliding stuff in amazon probably hasn't helped

13:16 commoditizing that and so that kind of leads to an interesting question is just um

13:22 you know what um you know what do you think are the biggest challenges i guess in your in

13:29 your industry at this time and what are the biggest threat and and how are you approaching that or

13:35 quite frankly does your whole industry need to reboot i know you're rebooting the company that's what we're talking about but

13:41 what what is are some of them still stuck in the old we call them bfs as big fat and slow so they're still like

13:47 making hardware versus you know changing yeah

13:52 there's no doubt that this has always been a very competitive industry and this is something an environment that we

13:58 are used to operating in and the key aspect to our ability to compete over

14:04 the years has been around our innovation and innovation is really a cultural thing that we

14:10 maintain true to our dna in the company and and to me innovation is about having

14:16 the courage to try new things and even to fail and giving juniper employees the belief that

14:23 failure is not is not a problem you know this is how we instill the spirit of risk-taking and

14:31 innovation the company that has worked so effectively for us over the years now your question was also touches on

14:38 um industry challenges well there are no shortage of industry challenges this is a challenging

14:45 industry at the end of the day probably one of the biggest challenges that we're facing right now is around semiconductor

14:51 shortages this is a very often talked about topic these days

14:57 it's not just affecting our industry it's affecting practically every industry today does it need a reboot i

15:04 mean absolutely today maybe around 90 of semiconductor supply comes from asia

15:12 maybe 90 comes from taiwan a small you know little island in in asia and i don't

15:18 think this is a healthy balance we need much more domestic manufacturing we need

15:24 more manufacturing in europe we need the government to pitch in because honestly this is a matter of

15:29 national competitiveness and security i am encouraged by some of the investments that u.s

15:37 companies like intel and global boundaries are making yeah until we're putting billions of

15:42 dollars to work and by the way as we speak uh the ceo is testifying on capitol hill exactly the pitch you just

15:49 gave it's not just a protectionist strategy it is a growth strategy but you're right they have put

15:55 billions of dollars to work and taking a hit on their income statement to get this going out or at least their balance

16:00 sheet i think it's the right thing to do it's the right decision to make by semiconductor fab companies in this

16:07 country and we fully support it yeah that's amazing um

16:12 and how do you kind of look at you know your co-creation efforts um

16:18 when you look ahead when you look at we like to call product road maps and boring stuff like that

16:23 um where do you see yourself and the the pretext under that is i've talked to

16:29 companies on this program over the last 10 years and some of them are hardware companies that could not figure out how

16:35 to change themselves into a service company i just mentioned one while ago who ended up acquiring some um bracken

16:42 darrell the ceo of logitech's been on they're still primarily hardware company doing extremely well of course a great

16:48 leader like you but yeah i guess two things how do you see the migration of of yourselves in the industry and you

16:55 see it more of a not just software as a services company with a balance of hardware but you see

17:01 going into other types of things in the future that maybe we haven't thought of yet yeah you know when i look at the history

17:08 of networking i've been in this industry now for 25 years there have been some key inflection points you know in the

17:14 80s there was around this unifying network protocol tcp that became the one language that that everything and

17:20 everyone spoke over the internet juniper bet its existence on that in the 90s

17:27 yeah and in the 90s it was around high performance silicon technology that allowed this the network to scale

17:33 juniper led this revolution i think if you look today the

17:38 key inflection is around software control points that remove the complexity out of

17:46 network operations and this is what we define as experience first networking now hardware remains very important at

17:53 the end of the day networking is inherently distributed uh high performance problem where you need

18:00 processors you need networking silicon to move traffic as fast as possible but

18:05 the promise of simplifying network of removing the obstacles of complexity

18:11 from network operations is first and foremost a software challenge and this is why

18:18 all of the software all the companies we've acquired over the last few years have been software most of our r d

18:24 innovation is in software today our software business is growing at a very

18:30 fast pace last year we hit over a billion dollars of sales achievement in

18:35 software so it is an incredibly important element of our

18:40 innovation um and it is the thing that's driving the biggest inflection point in

18:47 networking today which is around experience burst networking yeah yeah that is a big reboot by the way so

18:53 performance is good there you mentioned you just had a throwaway comment in there about how long you've been at the company i think your employee number was

19:00 it 32 or something or even less yeah is that right yes it's um that's amazing and that was uh

19:07 you've been there over four decades the way i measure because it's 90s you know and we're just getting into the 20th but

19:14 it's four decades so um what i guess a couple questions what the heck has your journey been like just personally forget

19:20 the business stuff here for a second you know coming at it from an engineering point of view you guys you and your competitors definitely are sales and

19:27 marketing culture because it is so competitive out there but i'm also hearing a a very strong engineering

19:33 innovation culture at your company as well absolutely you know i often get asked

19:38 why have i stuck around at what one company for so long it's a common question

19:46 it does come down to a couple of things uh first and foremost it's the people i

19:51 mean the opportunity to work and to learn from some of the smartest people in the industry and i've never in the 25

19:57 years i've been in this company felt like i'm the smartest person around i mean there's always

20:02 an endless amount of engaging conversations even as as a ceo

20:08 my favorite conversations are usually up on a whiteboard with a technologist solving a real customer problem

20:14 there's another aspect of this um i think that has kept me around is that is juniper is very much a purpose-driven

20:21 company you know we develop switching routing security products but at the end of the day we develop

20:27 something that's far bigger and more powerful than that what we do really matters what we do

20:33 helps our customers change the world if you think about just the last couple of years of the

20:38 pandemic i shudder to even imagine a world where we didn't have the network that kept us

20:44 connected to our families educating our children shopping

20:50 moving commerce along yeah this is all as a result of this incredible global

20:55 network and not just juniper juniper and our industry have helped create to me

21:01 i find this incredibly motivational and is why i've stuck around

21:06 in this industry and this company for as long as i have that's a really great way to reframe a

21:12 question by the way it's just like why have i stayed um i know a couple ceos that fire themselves once a year as part

21:18 of a practice it was popular in a book of years ago and and that refreshing that i've never seen them not come back

21:25 but whatever it's a tad self-serving but actually it's a brilliant way to get back and stuck back on on the you know

21:32 the actual purpose of what uh what's going on and and why you're there and you hit on another thing which is the

21:38 the internet those of us started out in the early days where we had to actually build it while it was going um because

21:43 it's such a point now i mentioned you know the games company when they build our own server rooms it costs about 50

21:50 million dollars to get a good startup going then um and infrastructure and all that cost and now i've got a slide where

21:55 i show every company i've been either ceo chairman of it gets down to almost like five grand right now with things like the aws and other things this guy

22:03 i've never seen a better time and a better platform for innovation and i know that's part of your your passion as

22:08 well as what has come out of the internet and that's so not necessarily in itself but all the capabilities that

22:14 we now take for granted um and and what's exciting more and there's

22:20 a question here is where's it going so things like the metaverse 3d modeling

22:25 lots of educational manufacturing design all types of applications there that are

22:30 out there even if it's just a boring zoom meeting but where do you see the potential of it going in the next 10 years

22:36 well i i touched on one area that it is where it's going and that is ai and ai as a

22:43 means to an end and that end i would summarize as the self-driving network in much the same way as we're moving to a

22:50 self-driving car we shouldn't underestimate just how big self-driving cars are going to be right

22:56 they're going to change the face of cities they're going to remove the need for parking structures they're going to save lives there is a similar need for

23:03 self-driving networks i mean another area that i can touch on that i think you know this industry is going is

23:09 around security as as proud as i am of the global network that we have built together as

23:16 an industry and here at juniper i think we also have to admit that it's a very connectedness

23:22 that we've created that has unleashed sort of unprecedented threats and unintended consequences we've for cyber

23:29 criminals removed uh time distance and identity constraints of doing bad stuff

23:35 so it's extremely important that as an industry we think about security embedded in every aspect of

23:43 what we do you know if in the past the constraints of building and expanding the network where

23:50 things like moore's law scalable protocols you know scalable silicon technologies i think the future

23:58 is around dealing with complexity which i already mentioned and then trust right and and that's where i

24:05 really take uh the security of the products that we developed uh really really importantly and you

24:12 know juniper's approach to security is not just about developing firewalls it is really around thinking about the

24:18 network as your threat vector as your threat posture to prevent the bad guys from

24:25 doing their dirty work because every ransomware attack every cyber security threat relies on the network to do bad

24:32 things well why don't the good guys rely on the network to detect those bad things and to prevent them where it

24:39 makes the most sense to prevent them exactly i do a lot of governance work with corporate boards and you know cyber

24:45 security was all the rage everyone got trained but it is still a defensive mode not offensive to your point that's they're

24:51 not they should be playing offensive just like they're being attacked attacked back and if anyone knew the amount of times the let's just say a

24:58 brand that you use maybe that makes food for you the number of times that that corporation was hacked on a daily basis

25:05 times it's thousands it's just normal normal course in space this curry is huge where else are you or the lack of

25:12 it right now but what are the big bets are you looking at yeah so ai security in the realm of

25:19 service providers uh 5g is obviously a

25:25 very big emerging opportunity and

25:30 within the sphere of 5g is this concept of network disaggregation

25:36 where what was once bought and deployed in these vertically

25:42 integrated systems especially for the radio access networks are now

25:48 slowly becoming disaggregated into separate components and separation of

25:53 software and hardware and the beautiful thing about this is juniper has nothing to lose here you

25:59 know we are again a challenger in this space and have an opportunity to

26:05 um deploy and to offer our customers a very disruptive economically compelling

26:12 solution for disaggregated ran or openram so we have

26:17 placed our bets we're working on technologies in something called a rick or a radio intelligence controller that

26:25 adds software intelligence to an end-to-end disaggregated open rand network we're

26:31 honestly seeing some really promising early signs in terms of interest for our

26:37 customers interest in the outcomes that we can deliver in terms of greater agility and

26:42 better economics of deploying these 5g networks worldwide something i'm super excited about

26:48 yeah there's 6g coming um but is there a trend to um

26:54 for corporates let's just say to bring stuff back on premise now that they've gone crazy with the cloud

27:00 i just i've heard in some circles that some people are just nervous about everything being out there yeah it just

27:06 seems like the train has left the station but what would say you well i would not bet against

27:12 public cloud uh i think every public cloud company is reporting

27:18 you know unbelievable growth right now and that is indicative of the direction for the market however

27:24 i do believe a few years ago there was almost this belief that everything was

27:30 going to move to the public cloud all workloads all applications i think today

27:37 people are far more realistic about what goes into a public cloud what stays in

27:43 um in a private cloud or a private data center that because that's just what makes the

27:48 most economic sense the challenge that we faced when we talk to customers around how

27:54 they balance those two things public versus private is they're just very worried that for the applications and

28:01 workloads they want to keep on private cloud they don't believe that they have the core competence the skill set

28:08 to manage a private cloud data center as effectively and efficiently as a

28:14 hyperscale public cloud provider yeah they've lost their muscle memory exactly

28:19 but this is where our our message our pitch is very simple you don't have to be an expert leave it to the robots it's

28:26 almost like a no close almost like no code low code kind of thing it's exactly right i mean that's right

28:33 abstra which is a company and a team and a technology that we acquired just um a year ago

28:39 in fact has invented this concept of intent-based networking which essentially means

28:46 it will simplify the act of operating networks to the point where it's essentially like you know using a public

28:53 cloud provider so there is really no excuse for any company out there that's thinking about

28:58 balancing workloads between public and private to not do that with their the technology exists and able to

29:04 enable them to do that as cost effectively and efficiently as possible yeah makes sense

29:11 i really want to thank you for for joining us today it's been a great discussion any final uh words of wisdom

29:16 for uh aspiring entrepreneurs out there or maybe engineers that we're employing number 32 about how they can like value

29:22 your footsteps and become a ceo someday it's a good question you know

29:28 maybe the one message i have for those aspiring ceos is that you don't need to jump around to move ahead in your career

29:34 i mean just look at me i've been at this company for 25 years i've done many different jobs at this company i will

29:39 admit i never had my eyes set on the ceo chair but that doesn't mean i wasn't ambitious uh and i didn't have this sort

29:47 of five-year 10-year plan but every i just worked super hard at whatever task

29:52 was given to me built a reputation of delivering and whenever that next opportunity

29:57 inevitably showed up i jumped on it i you know even if i didn't feel i was 100 ready i would go for it and that recipe

30:05 worked for me and it might actually work for those that are aspiring um for the corner office as well

30:11 i love that no matter what your next goal is but jumping into something you're not ready for is the most important thing you can do don't wait

30:17 till you're ready i love it thanks for joining us for me it's been great to see you again you've been

30:22 listening to him he's the ceo of juniper networks this is dean tobias with the reboot

30:27 chronicles we want to thank you for joining us today and we will see you soon