From architect to AI exec: Scaling leadership, maintaining credibility & effective communication to non-technical execs w/ Prashant Ramarao #222

We're doing a special in episode feature on the future of AI powered incident management with our friends and sponsor, X Matter. People as a primary integration layer is really fragile. With multiple people and all of that coordination, you become slower to find the root cause. The slower you find the root cause, you then don't know what action you need to take to resolve it. Getting to that fast is the goal.

Later in the episode, Mike Bennett, who leads the engineering team at X Matters, shares why human-driven coordination creates outage risk and how AI-powered orchestration can dramatically accelerate your path from event to resolution. That is a lot of hype. Sometimes when you hear stuff in in podcasts or in newsletters and so on, there is a fair bit of hype and you don't want to get swept away by that. And if you do

you'll you'll you'll probably lose your technical credibility. So I I think that is what I try to do or gain by doing these experiments myself. Because this stuff is so new, I want a first hand, you know, experience and exposure to this so that I know when somebody comes and tries to sell me something on a, you know, cool new project that we should be doing. I I can kind of argue and and kind of ask the right questions.

Hello and welcome to the Engineering Leadership Podcast brought to you by ELC, the engineering leadership community. I'm Jerry Lee, founder of ELC. And I'm Patrick Gallagher, and we're your host. Our show shares the most critical perspectives, habits, and examples of great software engineering leaders to help evolve leadership in the tech industry. What does it take to go from a deeply technical IC to leading over a hundred engineers?

How do you maintain your technical credibility and effectively communicate to non-technical executive stakeholders? In this episode, we are joined by Prashant Ramarau, SVP of Engineering and Head of AI at Yahoo to explore how he made that leap and the mindset.

that shifts that helped him scale. So in our conversation, we get into how to lead the senior most engineers communicating with GMs and non-technical executives, the reality of putting AI into product right now. Plus we deconstruct some of Prashant's favorite personal AI projects. and how those projects enhance his technical credibility and leadership skills. Let me introduce you to Prashan.

He's a deeply technical executive who still rolls up his sleeves. He spent his career at the intersection of AI, large-scale systems, and engineering leadership, defining strategies that drive real business outcomes. He's known for building high-performing teams.

challenging the status quo and staying hands-on even while leading a hundred plus engineers. Outside of his work, he is a self-proclaimed pod a holic who loves hiking Bay Area trails with an episode in his ears. If you're navigating the shift, From builder to strategic leader, or you're figuring out how to lead AI teams, this one is packed with insights. Enjoy our conversation with Prashant Ramarau.

Prashaun, welcome to the show. Thanks so much for joining us. I'm really excited for you to be here. It's great to have you. How are you? Doing great, doing great. It's it's you know, finally almost summerish out here and uh you know, we just got back from uh uh spring break trip to Japan. Oh my gosh. Uh so fantastic. That that must have been a phenomenal time with the family. Yeah, yeah. I'm like already thinking my next trip there. I love it.

I'm really excited that we could have this time together. For the people listening in, Prashawn and I have known each other for a couple of years now. Um we've had a couple dinners together. We've delved into a bunch of different topics around things like the future of AI and a lot of other different leadership topics at some of these different events.

And so Person, I'm excited for this because we get to spend a little more time diving into you, your story, the lessons that have helped shape you as a leader and some of the different moments that have made a big difference for you. So

So I'm really excited to have you have you here because I'm looking forward to this. We've had a chance to get to know each other over some of these dinners, but um this is a really fun experience for me. I'm excited to have you. Yeah, this is the first one-on-one meeting, I guess. And I'm excited to be here as well. I love it. So we want to talk about a couple different things. So how to be successful with people that are maybe less technical than you is one thing we talked about.

How to lead at scale or how to scale your leadership is another. And then maintaining your technical credibility. And you've had some pretty exciting learning journey experiences recently with with maintaining your technical credibility. So you were sharing with me a little bit about how you had this rapid transformation from individual contributor to leading over a hundred engineers where you had to learn all of those things and all of those things were sort of at play.

And so maybe I figured where we could start is with some of that story.

So can you bring us into the story behind that rapid leadership evolution? So what happened and then how did you get to where you're at now? Yeah, sure. So for most of my career, uh I was an IC and uh maybe about Eight or nine years ago I started managing people. Uh it all started with managing a small team of really senior, you know, distinguished engineers about six to ten. And then over time

It grew. It grew to actually it didn't grow straight straight from like six to hundred. It went from, you know, six to sixty and then we did a lot of hiring and it scaled it up to one hundred and twenty or so. But the interesting backstory there is You know, and and this is all you know retrospective thinking on what happened, right? Like, you know, so what uh uh what happens at the Yahoo uh I I've noticed is we rewrite our application stack every four years.

This rewrite happens for variety of reasons. Sometimes it's business reasons, sometimes it's You know, technology reasons and and so on, right? Like you know, mobile move to mobile, move to cloud, move to AI and and so on. The previous iteration of this re architecture was uh roughly around twenty eighteen and I was a chief architect and my boss at that time came and said, Hey, what should we be looking at as investment areas?

And at that time, this was twenty eighteen, so what, seven years ago, I was thinking about where is the industry shifting and it occurred to me that machine learning at that time is what we called it. Uh and deep learning is is something that we need to get really good at and get more people involved in it.

I pitch that idea, got a bunch of funding for training and got experts to come in and present and we trained up close to uh I wanna say hundred plus people in the organization, uh who went on to build products using machine learning. And then finally they were looking for somebody to lead

the AI team. And guess what? Because I had pushed for some of this early on, they said, Hey, maybe you should do it. It just happened and I was so excited to take on that role because of my passion for that area

I have a couple follow-up questions here. So what you sort of mentioned is you you'd spent most of your careers in IC, then sort of transition into management. And it sounds like the the first sort of group that you were you were working with in was the distinguished engineer group, which in my mind is like very senior, very intelligent, like very capable.

So what was it like for for that to be sort of the the earliest management experience to jump into sort of that distinguished engineer level as like the the first sort of management team you're working with? I I think for somebody who comes in late into management, having a senior set of people to manage. is much easier, I would say. They are very independent, they know what they need to do. Uh but oftentimes where they need guidance is coaching them on things like soft skills.

Right. Are they they are experts in their domain and and I had people uh in very different domains. Some were experts in mobile, some were experts in infrastructure, other people in machine learning. So it was a very diverse set of people, but where I kind of stepped in was mostly on helping them how to navigate, how to get their sort of ideas sold to other leaders and so on. So it was less

was you know direct people management, uh, but more like guiding them and they knew what to do after that. So this was the perfect way to transition into it. And I should give credit to my boss at that time who kind of recognized that. Uh yeah, it does it does it def it definitely sounds like that's a really fun group to jump into because it's really more how do we help unlock like

the vision of what we're trying to do together versus like making more maybe granular technical decisions and like maybe technical coaching and so I think that kind of sounds fun. So it's like there's velocity, there's direction. That's really cool.

So for you like so you mentioned like a lot of the role that you played was was more guidance and coaching and and helping people sell ideas. What would you say is bit was like the mindset shift you experienced, like moving from a more extremely technical IC role to then starting to lead

Starting with six and then expanding to sixty, then expanding to a hundred plus. Like what are some of the mindset shifts that occurred for you making that technical IC shift to to start to lead different groups? Right. funny thing is at at some point in my career I used to consider becoming a manager, like, you know, going to the dark side. But as you grow to into senior roles

you kind of start seeing and are exposed to what leaders do. And and I think that kind of was the first thing that told me that yes, this is a good thing and a good direction to go. I love to learn new things. I I always am on the lookout for what is the next thing I wanna get better at. And I saw this as an opportunity, right? Like, you know, leading a larger organization needs different set of skills. So so I took that as a learning opportunity.

In addition to that, I did have a really strong peer support group. Other people who had been leading large teams. you know, other VPs and S VPs in the organization I could uh lean on if I had any question at any time. So yeah, so it it it turned out to be a great learning experience and I I thoroughly enjoyed uh doing that. What was a moment where you realized maybe you had to start to focus on other stuff?

skills. So maybe a moment where you realize like, okay, I've been a a very senior technical IC and now I'm shifting to optimize for different capabilities. Like was there a moment for you where you realize, okay, cool. Like here's where the growth is, like or here's here's where I can really invest in that skill set.

I kind of knew all along that leading a team needs a different set of focus. Like, you know, firstly you start thinking more about people in your organization, right? Like, you know, how do you make them succeed, right? And and really kind of uh start having more deeper career conversations with people and those kind of things you never have to do if when you're an IC. But as I started doing some of that, I I I realized that my kind of mindset has always been a coach And even as an IC I was

the chief architect. So I was working with like, you know, roughly like six to seven hundred engineers and they would come and ask me questions on like, you know, how do I get promoted? What should I do? You know, and and those kind of things. So I had done that. as part of my job and not just as something on the side, uh, was was a nice way to transition into it.

Well, as you share that story, I also start to reflect on, you know, oftentimes like the IC track and the management track sort of get conflated as separate pathways. But really when you think about like your impact as a chief architect, like if you're mentoring and supporting the growth of six to seven hundred engineers, like that is also very much like a similar skill set for folks on like the quote unquote leadership management and management track.

And and the interesting thing is at least uh the way we used to run uh a lot of our things like performance cycles and all that, I would be involved in these discussions. Right. So although I didn't have to write ERN performance reviews for uh all these people, but I had to provide input, right? Because oftentimes I was the tiebreaker.

Because there are like, you know, two V Psy you know, asking questions and they would then look at me and say, What what hey Prashan, what do you think like about this person? Should this person get promoted or not? So that is really important in organizations where you have really senior ICs. Get them a seat at the table. Mm-hmm. Along with the other leaders. Yeah.

So I think what's cool about this transition is like in a lot of ways like there were sort of these like informal, like really large scale leadership roles that you were playing. And then now you sort of move into more of like a functional responsibility of of managing different teams and groups. And so I I'm curious, like in that switch or in that transition, were there areas where you experienced friction in terms of your own leadership or challenges?

With starting to play more of a formal or functional leader in some of these different roles? Like what where was where was some of the friction that you experienced? I would not say it's necessarily a friction, but it's just a different set of things you need to think about. Let's say you have

ten senior people, you can pretty much kind of figure out what needs to be done on a personal one on one basis, right? Like that that works pretty well. But once you are uh leading a much larger team, then you d need to start thinking about

Hey, now I need to have a quarterly w uh all hands, right? And then if you do have an all hands, how do you make it interesting? How do you get people to attend it and get value out of it? You need to start thinking about how do you communicate And how do you get it? information, you know, into the engineer's hand so that they can make decisions themselves. So

So the whole communication uh aspect is one. Of course there are other things like, you know, you start getting involved more in like budget planning and uh thinking about, you know, bonuses. and all that. So I th I think you just need to exercise a different muscle. Mm-hmm. And and many times you can lean on leaders who are reporting into you to kind of give you a lot of this information. But yeah, I I think there are a different set of things you need to start focusing on.

The moment you described earlier when you were talking about like the the person approached you as Asking, you know, what should we be looking at for future investment areas? Do you feel like that was a significant moment that put you in a place of more strategic thinking and contributing in that type of way? Or were there other moments where you were starting to sort of practice that more strategic, higher level? Leadership role.

Yeah, so I would say that like you know, as I said earlier, we at regular intervals would look at our application stack and say, Hey, what we have is this? Is this good? Or should we be doing something different? Right? So that was a fairly regular exercise, I would say. And and sometimes that exercise happens because of industry trends, you know. I don't know, I if I think back twelve years it was Stuff like CI C D was a big thing, right? Like, you know, get on to that bandwagon. So we always

This one and as I said in hindsight was a pretty critical thing, mostly in my career I would say. Uh and and also for the company, right? Like, you know, getting more plugged into the machine learning AI cycle early on was was a big

We're taking a quick break for a special feature on the future of AI powered incident management with our friends and sponsor X Matter. Mike Bennett, who leads the engineering team at X-Matters, shares why human-driven coordination creates outage risk and how AI-powered orchestration can dramatically accelerate your path from event to resolution. We're the ones that are correlating the alerts across the platform.

We're the ones that have to remember that a similar issue happened six months ago and this is what we did about it. We're the ones that have to figure out this is a symptom in service A. But it has a dependency in service B that we need to know what that dependency is and how that could impact this thing. We decide on who is going to be page based on some informal knowledge.

It's it's not scalable. I mean it all of that works in a in a very small scale environment. But as as systems grow, as teams grow, people as a primary integration layer is really fragile. So the outage risk is with with multiple people and all of that coordination, it you become slower to find the root cause. The slower you find the root cause, you then don't know what action you need to take.

to resolve it. The risk there is not knowing immediately what the problem is, so you don't know what the route for that mitigation is. With all of the information that is out there, getting to that fast is the key goal and is the key problem when you've when you're relying on people to do it. When a signal comes into X Matters, the first thing that you can do is based off of that signal, you can then make a call out to the right people.

Rydyn ni'n ymwneud â'r cydnol sy'n ymwneud â'r cydnol sy'n ymwneud â'r cydnol sy'n ymwneud â'r cydnol sy'n ymwneud â'r cydnol sy'n ymwneud â'r cydnol sy'n ymwneud â'r cydnol sy'n ymwneud â'r cydnol sy'n ymwneud â'r cydnol. From there, the incident commander can then use automations that are set up in the incident because it it automatically creates an incident for us.

It's linked to the ticket that generated the incident. And from there we can determine, okay, well I've seen I've seen this before because my incident suggestions is saying this looks similar to this incident you had last week. We've got built-in automations that can do stuff. So within an instant, you might have an automation that says automatically restart pods or automatically rollback services. Like I mentioned before, we can also do that as part of a response.

to the signal that comes out to say, okay, this has happened, do a rollback and I can just Touch my phone and go back to bed without even getting out of bed. all of the automation, the flexibility of the tool and all the the things that you can build in along with the data that you've got with the service catalogue, with your on call, with your who's on duties and get you to get the right people at the right time on the call if you need to get to a point where you're in a conference.

X Matters automates the entire incident lifecycle, taking you from initial event to final resolution. To see how their purpose-built AI slashes your resolution times and gives your team the context to stop disruptions before they start, head to xmatters.com. That's x-m-a-t-t-e-s.com.

Another element that we were talking about is this idea of the people you work with as you move into a more senior role start to change. And then maybe the audience that you're communicating with are are sort of less technical. Can you talk to us a little bit about what that shift has been like for you in terms of like the groups that you're working with?

And then we can start to deconstruct some of the the skills and the shifts to start to better effectively communicate with those groups. Right. So in my when I was leading that small team of architects, sort of every other week. There's a meeting with the CEO because

uh he wanted to understand what were we doing around machine learning, personalization. Are we kind of at the cutting edge or not? Where should the investment be and so on. So I used to go to those meetings and in those meetings I was mostly an observer. You know. Uh there were other people who did most of the talking and I saw how uh, you know, my boss who was

who was an E V P he kind of switched in a different gear. Like I could see that okay, you have to kind of, you know, talk to the CEO in a in a in a different way. And, you know, that was the first time I noticed that, you know, there is this different modes of And then when I started managing the larger AI engineering team, I was supporting six different GMs

And and I said, okay, the first thing I need to do is have regular one on ones with these GMs so that A, they know what we are doing for them and B, sort of understand what they are looking for, right? So I went into some of these meetings not really Understanding uh how to go about it. So my first couple of meetings with the GMs was strange I would say, right? Like, you know, I'm saying a bunch of things and I could see in the expression that like, you know, okay, why should I care about this?

kind of a thing from them. That's when I realized okay, no I need to kind of have a plan when I go and talk to uh these people. And it needs to be in a language that makes sense to them. And this is not talking down, right? This is like more talking in their language so that they understand why what you're doing is important for them. So that was

a lesson learned I would say. Uh but again going back to what I said earlier, I love to learn new things. So I I took this as a learning opportunity, right? I'm super comfortable talking to a technical person, but how do I kind of bring that to somebody who's Who doesn't think about technology day in and day out like me?

Well, I think that shift makes it more fun. 'Cause for me, sometimes whenever I'm learning something new, I'm like, Oh I'm like I fear comes in. I'm like, Oh man, like this is so hard. And the way you're talking about this shift is like, well, like, yeah, you know, look, let's approach it from learning something new and like anything else, like you can learn and and and approach it.

Like what did the learning pathway look like for you to to start to talk in terms of the GMs or these other other different roles? I did this course which was a masters in engineering leadership course where they brought in experts from the industry and talked to us about how to go about doing various things and one of those sessions was on just communication skills and like, you know, how to talk to leadership and so on. So that person talked about this thing called a contact plan.

was a phrase I had never heard before. And and I think it comes from the consulting side of the business, right? Like so I guess, like, you know, people, McKinsey or Alex Partners or any of these consulting firms probably use this all the time. So when I first heard that,

I said, okay, this is something maybe I should apply. Uh as part of the course, we got the opportunity to have a one-on-one conversation with the person who was teaching us a particular topic. So I talked to the person who had heard presented and I said, Okay, I'm gonna use this. And stay away from talking about matrix multiplication and how You know, Jen AI is

just an evolution of uh uh you know, machine learning and all that. So I kind of uh use that contact plan where you kind of go with a you know clear definition of What is the objective of the meeting? Have something that you teach them, have something that you want to learn from them and then finally sort of you know, ask them if there is specific things that they want out of you. That was uh sort of how I got more comfortable and have

you know, useful conversation with these people. So I've kind of used that. Now it's become second nature. I don't really think about all these things. I just like, you know, when I'm creating an agenda for a meeting, I just include all these uh different sections. I love the learning structure. And uh a person who really broke this down well uh many years ago through our some of our community events is Wade Chambers and he talks a lot about going from unconsciously incompetent

All the way to the progression of being unconsciously competent. And one of the different elements he talks about is like the role of deliberate practice. To help you go from like consciously incompetent to consciously competent is that you kind of have to deliberately practice and go through that exercise. But then what you're talking about is like then all of a sudden you've you've developed this like

like unconscious competence and mastery over it where it's starting to become second nature, but there's this whole learning cycle. And so as you're sharing this, I'm like, oh, this is the progression of learning. Like this is it. It's like you just have to take the thing that you've you've learned, apply it deliberately, and then all of a sudden it becomes easy or easier. Right.

So bring us back to the early days. Like how do you create your plan when you're you're what does it look like to create the contact plan? Like from a very like deliberate level. Now I it sounds like it's more second nature, but like what what does it look like to build your deliberate plan?

So, you know, you need to spend time creating your agenda and what you want to discuss upfront, right? There are some meetings I spend a lot of time preparing for and there are other meetings I just go in, right? Uh as an example if it's a little bit more than a little bit. uh one on one uh with somebody in my team, I take that as a time for them to talk to me about and bring, you know, their concerns or questions or

anything, right? So they should be creating the contact plan. But when I'm going to a meeting with, let's say, the CTO or or one of the GMs I kind of I create that contact plan and generally spend at least like, you know, ten, fifteen minutes before the meeting happens.

And sometimes I'll actually even start creating this because I know it's coming up in the next week or something. I'll just throw in things that I want to talk about and then remove things if there's not enough time. You have to put in some time early on. Can you walk us through maybe an example of what it looks like to build the plan and then have this like really effective communication with somebody who's like a CEO or or a GM?

So I think, you know, as I said, there are really four sections. Like first is what what is your objective? What do you want to get out of a meeting, right? It could be I want to get more funding to hire more people, right? Or it could be, okay, there is a challenge or there's an issue with some project and I want them to be aware of it. Right. So so you kind of have an objective set up. Then there is

uh what do you want to tell them or what do you want to teach them, right? So the teach section is generally about, you know, if if you want to say why you need to hire people. You tell them about projects you've done, how they've been successful, how they're contributing to some business metrics and uh so that they're aware, they connect the dot, right? Because most of these people have hundred different things going on in their lives.

then you uh learn what is their priority, what are they focused on, what do they want to kind of C uh you work on. So be very d you know, deliberate about all these different sections and go in and hopefully like you know most if if you have the structure it helps you sort of at least touch on all the points and not sort of spend all your thirty minutes Talking about something that doesn't really accomplish anything.

Well, I the way you break this down helps me really understand more of like what my role is in this conversation. I I run into issues where I have these meetings. I'm like, okay, I have a really clear objective. And then now I start to take on the role of like my job is to convince you that like this is the thing that needs to happen.

And I'm realizing as you're talking about this is that like that's a little more of a combative approach. And if I start to approach it from like who I am in that meeting, how I communicate then is so different and probably leads to a lot better outcome when it comes from a place of teaching and not telling. Right.

And then when you're talking about learning, then it's like, wow, you're already building in curiosity into the framework of this meeting, which then also helps you be more connected, have a clear understanding of their motivations, have a clearer understanding of their priorities.

The other story I want to talk about was the was the one where you learned about the the observer effect as a leader. Right. So this is this is an interesting uh story. So this is uh I can't remember I think I think this was uh probably around twenty seventeen, twenty eighteen, or maybe a little before that. Marissa Mayer was the CEO of Yahoo and we're presenting to her this fabulous CI C D system that

we were using which did thousand deploys a day, right? And we were very proud of it. Like it was a great system, you know, and and I'm talking about almost a decade ago and and we were releasing features at a breakneck pace.

And as the presentation's going on, I was in the room there was, you know, more senior people who were doing the presentation and, you know, David Philo, who's uh co founder of Yahoo, he was also standing in the back and he kind of tapped on somebody's shoulder and said, Hey, the site is not loading. Right. We said, okay. Let's go and take a look, what's going on? And very quickly we realized that there was a big outage. And and the meeting was stopped and immediately opened a bridge.

uh people jumped on to various uh channels, incident channels and and we were debugging. And this is all happening in that conference room. So Marissa was there all the time and she is on the bridge, in the meeting room, just taking notes. Right. And and this went on for at least an hour or two. Then she left. She had other things to do, but you know, she took bunch of notes, you know. Uh uh all of us saw that she's writing in in in her notepad something. And then she left.

And and we we were debugging and and it took us a long time for us to figure out what the problem was. Uh you know, the site came back up but root cause it took us almost up to midnight and and the meeting was maybe around ten o'clock in the morning or something. And we all went home happy that we solved the problem. And next day she sent a note to my manager uh about these are the things we need to fix.

Right. So the lesson that I learned out of that in our interaction was being aware. I think it was extremely smart of her to not start asking questions. Right. When there is this big outage going on, you really don't want to become the center of Attention, right? Like you know if obviously if she has a question everybody's gonna go and get her the answer.

And and she was very aware of it. And I think I've taken that away as an important learning, you know, so as you become more senior in an organization, you need to be aware of that. Right. Uh so that's the observer effect. Like you know, the minute you kind of go and ask a question. So even in a meeting if there are like twenty people, I personally try to not be the first one to talk.

Right. You want other peop to hear other people's opinion because the minute you say something, it'll just bias and tilt the conversation in a different direction.

So yeah, just being aware of that is one thing that I learned. The other interesting thing was like uh we all regrouped the next morning and said, okay, what do we need to do now? You know this is this is a bad incident. And there were like bunch of ideas thrown on the on a whiteboard, these are the things we should be looking at and we created this thing called SEAL team sale.

So essentially there were six different areas and we needed somebody to tackle each one of those and and they were looking for people to kind of jump in and work on this. And and a lot of these were very cross-functional, horizontal efforts. And everybody was wondering I I I saw people around wondering, should I take this on? This is not really my area of expertise. more like you know what is the right thing to do uh for the users and for the company.

you don't want to overamalyze it. Like I think I think it just instinctively for me it's made sense that you need to jump on this. This is the biggest problem to be solved. And it turned out to be, you know, raising my hand got recognized as something that

held the company and eventually got some awards and things like that. But anyway, so so that was sort of uh another learning which is like when somebody asks for help and there's a there's a big problem, just jump on it. You'll get all the support you need. Well, there's a couple things that occurred me as you're as you're sharing this story. So n number one, what in incredible way to internalize the power of the observer effect. Like this is a high stakes experience.

you have a couple like very legendary people in our industry in the room who are playing a very uh like a role of humility of like of like and conscious of their their power or their dynamic in the room and sort of withholding that and like letting the team execute noting improvements.

So I think one, what a powerful way to understand that. And then two, like what a cool way to start to surface. Like here is sort of career growth happening or like like Skill building happening live, where you're talking about like, you know, earlier on in this experience, you know, 10 years ago, like this was a moment to gain this really interesting cross-functional experience. And in the moment to be conscious of like, don't overanalyze, like, does this allow you to get to this?

this point and then give you this opportunity, but rather it was like, this is a big problem, jump in and then it continue to open up more experiences and more problems that you gotta solve later on. I think it's cool cool to sort of consciously see those th those different things.

Do you remember the feedback that Marissa was giving you in the team efforts? Was it more like process and like internal communication oriented or was it more like systems and like technical platform and like outage related feedback? I I think some of it was just a user experience thing. A as an example, I'll give you an example. When we had this problem, the user experience was they saw some error pages, right?

And there is no need to show an error page to an end user, right? Especially if if you are running, let's say, a news site. You you can kind of cache a copy and and show something that was a few minutes. old or stale. So so it was it was a lot of it was around that. And I think some of it maybe was also around incident management process, but it was mostly around how should the product behave when there are system failures.

Not everybody thinks about that. You know, I used to do a lot of design reviews and and we started adding that as a question, right? If your API fails, what does the UI look like? Right? You know, if if you're not able to talk as as an example, if you are in the mail team and and somebody opens the mail app, if it can't talk to the server, don't just show an error page. I mean that can happen. You still have the inbox and and you can show what you can show, right? So so think about those edge

scenarios and have everything all the way from design to product behavior, you know, to implementation. So think about these

error scenarios is is was was one of the takeaways which we were not doing a whole lot of before that. I love the idea of like how do you optimize like the user experience for an outage.

How do you think it from an experience design perspective? Like that's such a fun question and like totally off my radar because that's like totally edge case. But like outages can be transformational, especially in how you resolve them. And so like the user experience like can be such a big input for what is the resolution experience like? That's right. That's right. So so nowadays what happens is if you go to the as an example, the Yahoo homepage and if there is a glitch

you'll still see the stream. The newsfeed is always there. You might see a small uh note at the top which says this newsfeed was generated I don't know, five minutes ago because we keep a copy of the newsfeed. Uh and of course there are a lot of other challenges which is like, you know, the newsfeed is personalized, but if there is an outage you show a non personalised version. So

So it's not straightforward. That's why I think I think giving it thought is super important. And I I think people often overlook that. So I want to talk about maintaining technical credibility because I think a big worry uh for especially as folks are leading a more like a senior and technical like senior technical team, the the worry is like, okay, how do I earn their trust and respect?

And then the other hand is I know you've just been doing a ton of fun experimentation personally to like keep like your technical expertise up pretty high. So like part of it is just like I really am excited to talk about the projects you're working on. But so talk talk to us about that transition. Like when you started to take on these teams and your team starting to scale of of like more senior engineers.

What have you done to maintain technical credibility or how have you thought about that as a part of your your journey leading team? the domain experts that I have on m uh in my team, they are the domain experts. I I want them to have the expertise and let them teach me, right? So my focus

uh has often been to be more broad. And and if there is a need, I'm I'm kind of happy to go deep on in a specific area. So so one of the things I do is like, you know, I talk to people, let's say I talk to my data expert and ask him, okay, what what's what what are the big problems that you're seeing in your space and and kind of keep myself involved and learn. But I probably don't have the same depth as But as an example, we were looking to figure out what is the next

front end framework that we should be adopting, right? Uh so I do have a front end architect in the team and uh that person's the expert. But then I spent a fair bit of time. I knew that this was a big decision. This is gonna impact uh a lot of teams, I wanted to go deeper and and spend more time on uh learning the things. So in in those cases I kind of take a deeper dive and and will carve out time to learn.

Do you on the other side of that like'cause I'm I think what I'm trying to understand is like the decision of where to go deep and where where not to. And so is there is there something like a decision maybe you made where it's like, okay, I don't I I maybe I'll learn a little bit, but I won't go as deep. Is there a certain decision maybe where you pulled back?

how deep you went in terms of learning learning that element.'Cause it sounds like, Okay, cool. If it's a big decision, go deep. But what would be like an example of maybe the uh opposite of that where it's like, okay, maybe do a little bit of research, but like trust your expert? Um, I would say that

Uh, for example, we were looking at what next around observability. Mm-hmm. That that's an area which is actually fairly complex and and to get a really good understanding of How these systems are built and what platform to choose and so on.

really an expert. Like, you know, this person contributed to open source projects and and they really knew the details. So I said, Okay, I'm gonna make them the point person and check in check in periodically because they had the expertise. So I don't know whether there's a clear decision framework per se, but I kind of have a sense for which decisions are big and more impactful and and need me to be more plugged in versus When it's not needed.

I know right now you've been actively engaged in a bunch of different personal technical projects. with AI, LLMs. And so I wonder, Ivan, what are your favorite projects right now? But also why is this like experimentation important for your role? And like how does it how does like this level of experimentation like benefit your growth as a leader or like your impact as a leader? So I would say uh there are a couple of things here. One is, you know, obviously AI is so huge.

Right. It's extremely important for what any company is doing at this point. That's one reason why I think I want to be deeply plugged in. And of course, as I said, uh almost seven or eight years ago when I did all those deep learning specialization courses on Coursera and and whatnot and really learnt all the inner details of uh of how machine learning and deep learning work. So I want to be kind of in the middle of it because I think I think it's transformation.

So it's important for I would say anybody in the industry, if you can and if you even if you Can't you should be doing and spending more time. The second thing is the way I'm kind of spending most of my time is in two buckets. Right. One is just straight up using, you know, Chat GPT or Claude or any of these chat bots and I I I use all of them. Purely for analysis of material. So as an example, one of the things I was doing was in our local school district

there was a budget shortfall. And people did not understand what is going on. I mean the the school district was publishing tons and tons of you know, they they are like, you know, in three hour meetings and and pres have presentations, documents, spreadsheets. Parents don't have the time to go and like, you know, dig into this and and understand what is going on. Uh so I said, Hey, why don't I go and play with Notebook Alum? It gives me the opportunity to go and play with something new and be

Tells me what is going on. So I went and uploaded all the material that I could get my hands on and and put it in Notebook Allah and created this. you know, ten, fifteen minute audio summary which which explained what what the issue was, you know, and and sent it out to the parents group. Right? Like Hey guys, this is the problem. You know, uh they were like all the classic problems. We pay so much in property tax. Why is that not enough to fund the schools?

I I learnt a lot in that process too. I had no idea how the school funding uh system works in California now have a very good understanding, but so so just as an idea, like you know, so so those are like very practical things are going on. I know I did another experiment with You know, Apple iOS Health Data Downloaded the whole thing. I kind of learned how the whole Apple Health Data sits in XML files and there's a gigabyte in and maybe the size is because

I've been kind of pumping all my data for a very long time but and and then an analyzing it and and there were some interesting learnings there too. I mean I was trying to see if I was getting enough sleep because that was one of the things my objective was to get enough sleep in twenty twenty four. twenty twenty five is more about eating healthy, but yeah, it's twenty twenty four. Uh so I I was trying to do some analysis and I realized that, you know, actually L L Ms are not great with data.

Right. Uh uh so it was giving me all kinds of and a lot of hallucination. And then I realized that the real purpose of that exercise and and the learning was not about my, you know, sleep patterns, but it was more around what are LLMs good at and not good at. So that kind of I think was the interesting uh learning.

This is so great. I love the everything about those experiments. Just a quick comment. Those are those are absolutely a ton of fun. Yeah. It's it's very practical, right? The other bucket is essentially just using cursor or wind surf to write code. In the past I would, you know, sit and if I have to write code for some just fun project it would take fair bit of time. the turnaround time for a project would be like a month because I'm spending like, you know, half an hour

uh over the weekend or something like that. But now with windsurf and I can finish that whole thing in like you know thirty minutes. Right? So so that is the other bucket. So that has enabled me to do other fun projects. Uh just because I can write code much faster with these tools.

Well, that that acceleration is so real. I mean, we're we just did an event last week in in the South Bay and it was on like the blurring lines between engineering product and design. And people are sharing these experiences where like their PM is mocking up something in a couple hours. And then they're launching a feature that would have taken a quarter in a week or two. It starts with first being presented a working prototype.

And then engineering and product and design, then working together to accelerate and bring that together. So, but just for you, like to take a project that would have taken a month and accelerate to thirty minutes, like that's awesome.

So to connect the dots to to technical credibility. So these personal projects, like is there maybe a short summary for how how this is relevant and impactful to your technical credibility with the things that you're doing? Yeah, I think one thing that I would say is It keeps me grounded. There is a lot of hype. Sometimes when you hear stuff in in podcasts or in newsletters and so on, there is a fair bit of hype and you don't wanna get swept away by that. And if you do

you'll you'll you'll probably lose your technical credibility. So I I think that is what I try to do or gain by doing these experiments myself. Because this stuff is so new, I want a first hand you know experience and exposure to this so that I know when somebody comes and tries to sell me something on a you know cool new project that we should be doing. I I can kind of argue and and kind of ask the right questions.

The first hand exposure helps you know the capabilities, it keeps you grounded in reality. Right. I love it. Prashant, we've got a couple of rapid fire questions. Sure. Um if you're ready to jump into those. So first question, what are you reading or listening to right now? So I I mean, I don't know if this is a term, but I'm a part of right. I listen I listen to so many podcasts.

I think uh I probably clock more than thousand hours of podcasts a year and I know the math geniuses will do the math and say, How can that be possible? I do listen at like two X speed. That's right. That's right. That's the way and there is a whole story and science behind why that is actually a good thing. The weird thing is I think I sometimes don't know what people sound like in real'cause I always listen to them and I'm not sure.

uh two X the speed. So yeah, I'm I'm kind of um I spend a lot of time uh listening to podcasts. The other interesting book that I'm reading slash listening to is Good Strategy, Bad Strategy by Richard Rimmel. It's uh it's a classic, you know, how do you identify when you look at a strategy statement from someone, is this a is that a good one or not is what I want to kind of create a filter for. So I'm kinda spending time on that.

Love that. Well you have to leave us with at least one pod recommendation. Oh w um of course ELC has been uh and I have to say that like that's something that I very regularly listen to. Other one uh is Dwarkesh. So I know that's a more technical podcast, but yeah, that's that's on my feed. That's great. Next question. What tool or methodology has had a big impact on you?

So I kind of you know, in one of these podcasts I listened I was somebody was talking about it and this is not an original one. What I have started doing is if I'm commuting or I'm driving and have A good half an hour plus in the car. I just turn on the chat GPT voice mode. And, you know, ask it to teach me something new. Uh or if I have questions about some s particular topic, I just kind of have a conversation and I learn about it.

I didn't know much about how car leasing worked and why do companies want to lease cars? What so so yeah, I learnt all about it on my drive to dinner one of ¡Suscríbete al canal! Self directed learning or uh creating your own rabbit holes. I I love it.

I've fallen down the the car leasing YouTube rabbit hole for a while. There's like a there's like a YouTube channel that does like all this role-playing of like here's what car dealers are trying to do in negotiation with you. And I'm like I'm like, I'm not buying a car, but I'm just like absolutely fascinated by it. So I was like, Yeah, I I love the rabbit hole effect.

This next question is about trends. So trends you're seeing or following that's interesting or hasn't hit the mainstream yet. But I know you've also been thinking about like where the wins are within AI. So we can answer that either direction. So if

there's a trend that's interesting or hasn't hit the mainstream yet, we'd love to hear it. If there's a different one, we'll love love to share the the alternative answer to this. Right. I I'll I'll say that like, you know, this this came up in a conversation with somebody who worked in the healthcare uh health tech space and when they first mentioned it I was okay this is interesting. It's called ambient AI. Essentially it is

AI everywhere you are. As an example, in I think in the the healthcare space, it's essentially the doctor's office. uh where you know the doctor usually if you go in opens the computer, enters a bunch of things, this that in ambient you know with ambient AI they just

come and talk to you. They're not distracted by hey, I need to log this or enter this in some system and all that. It gets just taken care of, like you know, takes care of the process, takes care of the workflow, takes care of prescription, whatever. Now that's in the doctor's office, but I think, you know, this Could be

something that can be even outside that is the thinking here. So I think there could be a trend that that ambient AI could become a big thing. Of course there are all kinds of privacy questions and whatnot so needs to be done carefully. But I think given its capabilities that

That's possibly one of the directions where things could go. Interesting. All right. Last question, Prashan. Is there a quote or mantra you've lived by or a quote that's resonating with you right now? quote that I really you know, really live by and believe in is uh by uh Rear Admiral Grace Hopper which says the most damaging phrase in the language is

It's always been done that way. I think we need to keep challenging status quo. Otherwise uh otherwise we are doomed. So so I I love to question anything that has been done. And if somebody comes and says this is What we have done? No. Because the circumstances change, right? Like there is technology shifts, there are business uh shifts and and so on. So I think I think there is always an opportunity for uh everything to be questioned.

Amazing. Prashant, thank you so much for spending time with us. It's been awesome for you and I to be able to just sit down, hang out one on one. Um so thank you for sharing your stories with us. Absolutely. Thank you for your time. If you're listening to this and you're wondering, how can I connect with other engineering leaders in my city? Pull up your phone right now and go to elc.community, click our

chapters page. You can see that on the menu on the left. Find your local chapter and click join. We're hosting virtual and in-person events all the time. And this is the best way to help you get involved, expand your network in your city, and support your leadership and career growth.

So pull up your phone, head to elc.community, join your local chapter, and get involved. A huge thank you to all of our local leaders who make community happen, and thank you for listening to the Engineering Leadership Podcast.