Name: Women in AI: Leading Enterprise AI: From Experimentation to Meaningful Adoption
Uploaded: 2026-10-23T17:31:12.550Z
Duration: 43 min 52 s
Description: Women in AI: Leading Enterprise AI: From Experimentation to Meaningful Adoption

Transcript for "Women in AI: Leading Enterprise AI: From Experimentation to Meaningful Adoption": Hello, everyone. We're gonna give it just a couple of minutes for, everyone to kinda get logged on if you're joining already, and then I'll go ahead and kick us off. Oh, Sarah from Croatia. Hello. One of my favorite trips that I've ever taken. I still need to make it out there, and I've heard so many great things about Croatia. Anjali, it is beautiful. I've heard so... And I've seen pictures with my friends who visited. So, Sarah, we are coming. San Jose, Austin. Thank you all for joining. New York. Yeah. I didn't know honey is in New York. I never knew that. I have to tell you, I work with Honey, and I did not know this. I'm always better in Santa Clara. Actually, San Francisco, you know, through the lens. Yeah. Yeah. For those just hopping on, we're gonna give it one more minute for folks to roll in, and then we'll go ahead and get started. Calgary, another lovely place. Britney, I also sit in Denver. Should be our last day of lovely weather before a string of rain in Denver. What... We've been getting some rain in the Bay Area, but it's just getting started. I know rain is important. We need rain, but I do wish for fall, clear breezy fall. Okay. We're two minutes in. So I'm gonna go ahead and jump in. I want to welcome everyone to our first women in a I... AI event hosted by Glean, and also welcome someone who has quickly become one of my favorite people, Manjari Agochiya. Manjari, I would love for you to kind of give a quick... Just two minute introduction to you, where you are now. I think, brief history of where you've been, that's actually one of the most interesting things for me. And then I'll kinda jump into questions. Yeah. First of all, I wanna thank Glean, and I wanna thank Jessica, Lowitt, and Manjari for having me here today. I I really appreciate this opportunity, and thank you all for joining in. You know, this is very close to my heart, and that's why when Manjari reached out to me, I was like, I would love to be, you know, on this event talking to all of you. So first of all, thanks for joining and listening to us. But a little bit about me, I started my life in this country as an immigrant and went through all the immigrant challenges that a lot of people... A lot of you must have gone through. You know? But, as a... In my background, I did my bachelor's in psychology, sociology, and my aim was to become, you know, an IS officer in India. And then it quickly changed to, hey. I wanna do an MBA because I think that is more exciting. I wanna go to the corporate world. And then I sat for all India entrances, and I decided because computers was the in thing to do, and there was only a few colleges which are offering that. I got through my entrances, and I did a master's in computer applications from one of the top b schools in India. After doing that, I came to this country, and I thought life is gonna be hunky dory because I have gotten through the toughest part of getting through one of the best schools. And now, you know, I graduated one of the top students. Life should be great. But that is where my life started, and it was a learning experience right from that point onwards. You know, when you're in a school system, you feel rewarded for everything because you're getting good grades. You're, the reward system is very structured. But as you step into the open world, you realize, hey. You're not gonna get rewarded for everything good you're gonna do. Right? That's the biggest waking moment you have in life. And that's what happened to me too. I came to this country, and and things were not as expected. So I had to go through a lot. I did a lot of... Started a lot of businesses. I did a lot of projects. I started as a as a full stack engineer developer and then quickly realized I couldn't I couldn't spend my life sitting... Coding behind a computer. I needed more people interaction. That was me. So I moved into project management, program management, moved up, you know, doing different types. I was in the fintech industry, different industries, and just wanted to get a flavor. Learning is my passion. I love to learn. And from there on, I... It was time to have children. And when I decided... When me and my husband decided, hey. We wanted to have kids, and we had to go through other challenges. And finally, when we had them, I wanted to relish them. So I took a break from my career. I, enjoyed raising them, and then I came back. And when I came back, my first four day was in the AI space, and that was back, about six, seven years ago back in Wells Fargo. My... It was a huge program. We were, you know, we we actually went straight into building models and predictive models and analytical models, and and I didn't even know what a model meant at that time. I didn't know what AI meant. Forget about what models meant. You know? And so it was a very steep learning curve. And because the program was 500 stakeholders across 29 business functions, and the release was for all of Wells Fargo. So the impact matrix was very, very large. My yearly budget was over $80,000,000 at that time, so that was huge. That is huge. And and so that was my first foray into AI. And, and then from there on, I came to Uber. And in Uber, my life started with HR tech stack, and and then I defined... They... You know, we went... I went from... I was very fortunate because I went from one large program to the next. And then, then Uber was defining ads in a very different way, and so they had, like, it... I think it was 2,000,000,000 ARR or 1,000,000,000. There was, like, this big number they had and, you know, they had to reach it, and I actually led that program to fruition in seven months. With my success in that space, I was brought into cloud migration because that was the next thing that Uber was doing. And, we were basically transforming everything from on prem to cloud. And so I led that. I led four data centers to the cloud, with over 2,000 bespoke tech stacks. And some of these were, tech stacks from when Uber was just founded. So nobody... There's a lot of legacy stuff. Right? So nobody knew what was going on. And from there, I came into AI. Right? And then AI was the next big thing in 2023, and, my chief architect officer, Albert Greenberg, at the time said, Manjari, this is it. Do something with it. And I'm like, but what do you wanna... Want me to do with it? Nobody knew what AI was supposed to do. OpenAI was a baby at that time. They had just launched our GPT, and there was a lot of buzz, but there was no understanding of what it could do for anyone. Right? So that's where my journey truly started in GenAI. And it was... There was no looking back and, some of... And I'll... Jess, as we move along, I'll talk about some of the things we experience, and it's been fun, fun, fun. It's been one thing to the next to the next, and it's been learning all along. You know, the one thing I think is constant for me in my journey is learning. And today, four years... After four years at Uber, I've joined Marvell Semiconductor about three months ago. It's a semiconductor company, has been making the news, fairly... Largely in the last few months, since Nvidia and AMD have shown a lot of interest, and semiconductor and the chips, you know, are the the blood flow for all the AI transformation that we're doing. And I personally wanted to come into semiconductors because I wanted to learn what it was all about. Because I do believe that AI strategy is not about the horizontal. It's about the vertical. Eventually, the intelligent companies are gonna move from thinking just about the application layer to to actually think about strategy right from the application layer to the to the bottom most, which is your chips and how chips are gonna get optimized in the future for, for, you know, for proper AI transformation. I think that's where eventually it'll land, but, hey, that's what I feel, but who knows? It's a little. this is a little I love it. It's funny because, actually, the reason I had joined Glean was very similar. I knew AI was... I mean, everyone knew three years ago it was going to be important, but we didn't know why. Right? yeah. Yeah. And I took the path of saying, like, okay. I'm gonna force myself into an AI first company, so I have no choice but to learn. And I think that's, That's what it is. that is, yep, super exciting. I'd love to know... So you started as, you were coding product manager. Right? And you talked a little bit about your journey moving up and that people are asking you to solve problems that don't really have an answer. Right? Did you make a concerted effort to put your name out there as a problem solver, or did people just see that as a result of your work? I think it was more a result of my outcomes. I am very deliverable and outcome focused. I always start with the why, and, and most people don't. Right? Most projects come... Get thrown at you because somebody thinks it's important. But why it's important, people forget about it two months after they started it. Right? But if you keep that front and center and you keep the why in front of you, then the deliverables are obvious. And that's... I mean, all of us sitting on this call must have experienced it, right, at some point in time. And for me, that is important. I think I have went I went from one to the other, not because the... Of the absence of challenges or the absence of politics or the absence of difficulties. I went from one to the other because I never lost track of what I started the whole thing for. Yeah. Okay. Yep. I love that. Same. I think that resonates probably with a lot of folks on the call. For those who are listening in, there is a q and a option. We'll get to those questions later on, but if something comes up, please feel free to pop them in there. I'm actually... I wanna kind of pull that that thread there. When we think about AI experimentation compared to actual meaningful adoption, a lot of organizations get stuck. And I'd love from from your perspective, you've you've done this at Uber, and now you're doing it at Marvell Technology. Right? What really separates teams that are successful in making those transitions versus ones that kind of stagnate or plateau? I think it's very simple. Right? It's the willingness to change. It's the willingness to adopt. It's not being scared of what's gonna happen to my job. It's not being scared about what's gonna happen to x, y, and z. There's a lot of fear mongering in the industry today with AI. Right? There is a lot of opinions, and we are all... Because of social media, because of the exposure that we all have, we're all exposed to it, but one has to know what is right. Right? It it comes back to the basics. It's about outcomes again. If you have adoption, which has success, related to it, success metrics, defined, then you wouldn't get lost. Then you're fine. Then it's meaningful adoption. Right? If you if you tell your leaders that you're changing for x, y, and z reasons, and this... They align with you on those x, y, and z, reasons, then you're good to go. Then it is meaningful adoption. But the thing is there has to be alignment. You shouldn't be running on your own. You should have... You should always have some sort of an alignment where people on the top... And if that requires education, then that's where you start first. You gotta educate your leaders first. That is not the lost cause, actually. That is one of the most important missing links, in organizations where if leaders are not educated, then they don't know why you're doing what you're doing. They may just feel it's experimentation. Right? And. think about it if you're busy and if you have 15 meetings in a day, they probably have 35. So you have to understand that, that that is one of the key things you need to do first. You have to align with your leadership. You have to educate them. You have to make them understand why you're doing what you're doing, then it becomes more meaningful. Yeah. I think a lot... Especially now... Like, you think back a year and a half ago, everyone was talking about productivity with AI. And I think now that as a loose metric has really fallen flat. Right? And people care way more about what's moving business metrics, like things like revenue or cost. I'm curious when you're trying to teach leaders and c suites or VPs why this is important, what metrics are you leaning on usually? I always lean in on the workflow. So. what is your big metric? What is your company trying to do? What is your... You can... Uber was a product company. We had two apps. Right? The ride share, and and we have several verticals in Uber, but but it was a very different animal. Marvell Semiconductor is a semiconductor company. They produce chips, very different animal. But at the end of the day, what does my CEO want? Right? They want. a top line, and they want a bottom line. That's what the board wants. That's what the people want on the top want to see. So if you can show a road map which aligns to that story. Right? Hey. Whatever I'm doing at the foundational level will eventually have an impact on your key metric on the top. Now as easy as I am trying to say it, it is it is not easy. It is absolutely not easy to do this. This is a gargantuan task. Right? And it requires a lot of focus. It requires a lot of leadership alignment. It also requires peer alignment. It requires your team to understand what you're doing. So there's a lot of hard work in connecting the dots there. But if your vision is clear, then it's easy to translate that. And and I always hone in on that. My value metrics are always on the workflow. So it's it's not changed essentially from what we were doing earlier, right, when we had no AI. We were still doing this. We had KPIs. We had success metrics. Why are we not thinking about it now? What has changed? It's just a technological shift. Right? Yes. It is a different type of a technology. Yes. It has a different type of impact. So that understanding has to be there. What can this new technology do for me? But at the end of the day, your principles don't change. Your foundational stuff is still the same. How you calculate value does not change, believe it or not. ROI is still ROI. Yes. I love that. ROI is still ROI. This is something I say regularly to the the go to market team here. Yeah. Okay. I think this actually is a is a pretty good piggyback question to what you just said. Right? So in... We're in this rapidly evolving field. AI, every single week feels like a month. Every month can sometimes feel like a year. And as a woman, how have you stepped into this space, built credibility and influence, especially in moments where the technology and the strategy and the path forward is still in the process of taking shape? I'll tell you what I did at Uber because that was my first foray, into doing this. Right? So here's the thing. If we are scared, others are scared too. If we are ambiguous, others are ambiguous too. So one has to keep that in mind. Right? So what do you do when you have a very complex situation at hand? I always used to hear this from my parents. What do you do when you have to eat an elephant that's one bite at a time? Right? So you have to define what those bites look like, and that's what I did. So the first thing I did at Uber, right, was to identify who are these people who are super excited. The second biggest thing I did was, I convinced my leadership that there is value in this. The third biggest... And and these are big things, by the way. There's a lot of underlying work that I had to do to get my leadership on board. The third biggest thing was I got alignment. We... I I did not have a title in AI at that time. I did not have, a title which could draw influence at that or draw, like, an automatic nexus of people to me. No. But I made sense when I talk to people. And what that was was, hey. This is what we need to do. This is what our CTO wants. What do you think we should do? So I aligned who were the key people, security, legal, compliance. I brought all the business leaders. There was a platform, AI platform already in place. I brought him in the mix. So I got all these people on board, and I said, look. That was my AI ops, essentially. And I said, this is what we're doing. We need to self organize. So the first thing we did was from every vertical, which was important. Now this is not the business units. The business units were our consumers. Right? We were the central nexus, who was gonna drive the whole mission of Uber transforming forward. And I had some brilliant people there. Right? There was my head of legal over there. He gave me two people. He said, I'm gonna allocate these two people who are only gonna focus on... On AI contracts. Right? Then I had my strategic sourcing team. They did the same. So we had two to three people for every vertical, which was a key player in decision making. Right? So... And then we organized our processes really quickly. So, like, for for example, third party risk They came up with a brand new plan of how we wanna assess AI at Uber. And, therefore, we had a strike team of, I'm not kidding, just 10 people. Max, 10 people. And we ran a company of 52,000 employees, the e x t's and FTEs. And we were the only company, and I can say that very proudly because Harvard Business Review did a study with us, interviewed me, and they interviewed other FAANG companies. And they said, you're the only one who's been able to successfully do it in a centralized fashion, in a unified fashion. And and and we were able to actually get through a 168 AI intake reviews in a year, which was unheard of at that time. is unheard. of. Most companies were doing probably three, five, seven at the most, and they didn't know how to... And and a lot of them were doing it at a team level, at a level where leadership did not have visibility. In our case, we had... Our leadership had full visibility. Our... I had finance fully aligned. We had a centralized budget. So they knew where the money was going. There was no ambiguity. So it all depends on how you wanna do it. Right? Each one of you is sitting here is a keen learner. Right? You're all learners. You're all doing it in special ways. And I think this is how I did it, and this is how I found success. It's the AI in silos that causes such slow transformation. Exactly. Exactly. So when you get the silos out, that's that's the first thing I'm doing at Marvell. That's the first thing I did, clear all the bottlenecks. I don't want teams where I have people sitting on the top just relaying messages from here to there. I want. self organized teams. I want teams who have... Every person on my team is a leader. I I don't look as a develop... Look at a developer as just a developer. My developer should be able to tell me the full story. So that's how I look at people, and that's how I raise my teams. Oh, I love that. That is the, I think, the ideal state for everyone. And it's good that it is... Like, there is an example... Multiple examples now of you going to companies and making it possible, which is great. One of the things that I hear about a lot, especially in the AI space, is... And I have been subject to this too, is you're the only woman in the room, dealing with, you know, your opinions being the only different voice, and that leading to feeling sometimes imposter syndrome, sometimes, like, questioning if you're totally off base. I know if I've been there, definitely. How have you have you dealt with this in the past? And if so, how? On an everyday basis. And, and, yeah, I'm not alien to that feeling. Sometimes you wanna be in tears, you know, but you wanna hide the... I mean, I'm like, hey. Come on. I... I'm making sense. I've been in a room. I'll tell you... Give you guys an example. Here's what happened. I was talking to, somebody very senior at Uber, and there were four guys in the room and my... And there was me. And the the risk I raised was, are you thinking about the fact that tomorrow every SaaS product in the company is gonna be AI enabled? How are we going to think about that? And at that point, we were only sitting at a... In a space where I had, what, two enterprise tools at Uber. And, and so then these... Everybody in the room starts talking and say, oh, she's trying to say this. And then they're they're reeling my message to each other, and I'm like, I'm in the room. You know? I can tell you what it... It's a very subtle example of what happens. Right? And so how do you make your space is you always are... Should be fact driven. I was... I always made sure that I am the most educated in the room so that... And that required a lot of effort. And I always had my t's crossed and my i's dotted, and I still do. Nobody can challenge me on my facts because I do... I have the hardest working when it comes to my facts. And so if you have that, you will always have a voice because they also know at the end of the day that you are... They are going to have to listen to somebody who knows what they're talking about. And they can ignore you once. They can ignore you twice, but they cannot ignore you 10 times. And that is exactly what happened with me. Have I been subjected to this over and over again? Is it scary? Sometimes. Yes. It is. Is it intimidating? Absolutely. But, you know, that's when you know who you are, and that's when you know what you know. And you can just leave the room saying, hey. You guys, when you figure out your stuff, let me know, and I'll come back and talk to you. This this makes me feel so seen, Manjari. That's just exactly how how I feel like I also operate. Right? Know the data, know the facts, know you're. right, and let people continue to share, continue to say it, and let people come to the realization that you knew what you were talking about from the beginning. Absolutely. Okay. I'm gonna shift to a couple questions from the the group. The the first one that I wanna I wanna go through is, Melanie says, I agree with the need for leadership alignment. But in my experience, it's been that leadership is 1000% in on the need for transformation workflows with AI, and it's the day to day users that are stuck or not seeing the value. So I'd love to. hear your perspective on this challenge and how you've approached it. Yeah. I mean, that is also one of the biggest challenges people see, right, in due... Today's day and age. And there's nothing called perfection. Right? So in in a scenario like that, I would work with people who are excited. Work with the ones who wanna work with you. And remember, you know, you're not here to change the world. You're here to change the few who will go change another few and who will go change the world. So it... Sometimes we want to make greater changes than what we think, what what we possibly can. So you you know, it's it's an issue that you can accept and say, okay. I'm gonna pivot my approach and then work. That's what I did. Several times, I only work with the ones who are excited, and then the excitement flows. You know why? Because they go and talk about it to their colleagues, and they're like, do you know I attended this? And here's what I heard. Have you attended it? Have you tried this tool? A lot of it happens like that. Right? People may not listen to what you have to say, but they will listen to somebody they trust. And as long as you get to those few people who they trust, the trust circle starts to increase, and your impact area starts to increase. It may not be at the time and the pace that you wanted to go, but it'll eventually land there. But it's a risk and then it's something you have to accept. Hey. That's how it's gonna work in your organization. Yeah. Going and talking to people and understanding where the where the consistencies and the alignment lies across everyone that also moves up to that. that company level metric? Yeah. And, you know, sometimes you may have these people who are, I would say, who are blockers. So there are these... Some individuals who are blockers. Everybody has them. And it's not that those people think that they are blockers or they want to be blockers. They are just in a mindset where you... When they don't align with your approach, essentially, or they think they wanna do AI differently, or they think it's not necessary for them. Those are the cohorts I have personally gone and taught one on one. Right? Especially if that person is an influencer in the organization. Then you take the time to educate them one on one because for them, sitting in a public forum may be intimidating and you don't know it. So a lot of it is about psychology and organizational mindset, then it is about technological transformation. Right? Actually, this is so interesting because I think this is something that sets female women in AI apart from the general mindset of many... I'm not gonna blanket statement all men, but many men that I've worked with in AI, where if someone at the at the front lines is like, no. I'm I'm not gonna do this. It's, well, you're gonna do it anyway. Whereas we approach it as, well, help me understand. Yeah. Help me understand why you don't want to, where are we coming from, and then how do we get you there. I think that is one of our superpowers 100%. in AI as women for sure. 100%. And and I feel like we... Nobody can do it better than us. Absolutely. That's why the next one should be men in AI, not women in They're the ones who need the help. Okay. The next question, I'm gonna take from, I believe it's pronounced Manjari. Sorry if I pronounced that wrong. But with AI, Yeah. everyone is building. Everyone is vibe coding. Right? How are you making sure that these small side projects are actually building up, like, true ROI for the company? Very hard to do. Right? And and and, frankly, I mean, what I'm seeing... Actually, what you would see when you start transformation is that 80% of it, 90% is experimental. And that's okay. Right? That's okay. But now you want them to get to a place where they are deriving actual value. So then these people need coaching. Now you can do that through trainings. You can do that through one on ones. You can... Depending on the time and the resources you have. I... When I do value exercises, I like to do workshops with people, especially business domain workshops. And what I do in those things is I talk about, okay. This is fantastic. This is fantastic that you've built this, but tell me now, you know, what are you doing with this? What is the actual outcome with this, and how is it impacting your workflow? So then you are going to be... Once you have those in... Intimate conversations with business domains and with these teams, there will be value that comes out. Now whether that value is meaningful to your organization or not is a different question. Right? Then you can coach these teams to think around, maybe this is not the right thing to do. But to get to that point, you gotta get to know them very well. And you have to invest that time now whether you do it through your team, whether you have a... Most of us are lone rangers or has very small teams doing this. So, yes, I understand it's very hard. So how do you start there is you take your highest impact group, and you take your projects in the highest impact group, and you start working with those. Or the other option is you work with teams where you're seeing the highest token consumption. And you work with those because that's where you wanna do the spend gaps, and that's where you wanna educate people the most, that means. Because what if all the spend is frivolous? Right? It's. gonna cost somebody's job at the end of the day because finance doesn't look at it as value and meaning. They look at it as harsh numbers, and nobody likes those realities. And so sooner than later, it will have an impact. So before it has an impact, you have to catch it. And. my approach is always either the highest impacting domains or the highest token consumption domains. That's so interesting. The token consumption one is consumption one is something we hear a lot. I have some assumptions on what impact means, but from your perspective, what what would be an example or two of a high impact domain? Like, I can tell you, for Uber, it was engineering and. customer operations. Right? We had a very, very large team for customer operations there because, obviously, you know, we're a customer facing, product and app. So that is our bread and butter. So those are your high impact teams. Now at Marvell Technology, it... This is a purely engineering company, and engineering is my bread and butter here. My workflows. See, I think my leaders... And I think... And I'm not... This is, you know, I'm I'm three months in at Marvell Technology still, and we're still kind of forming a point of view here. But the leaders would not care about whether HR is transforming or not. It doesn't sound very good, but, yeah, that's the reality. Right? Let's let's talk facts here. So you... Maybe you find a good use case and you're gungo, and there's somebody in HR who's gungo about transforming. You support them, but they are not where you focus. You have to focus on what your leader wants to focus on, on what they think is important. And in this case, it's engineering. 80% of my token consumption sits in engineering today. So. I'll give you an example. We were trying to form a FDE cohort to support the nonengineering teams. And then when I saw the metrics, I went and I spoke to, my manager and I said, you know, I think our focus area has to shift. We need to shift to engineering rather than nonengineering because if 80% of my tokens are getting consumed by engineering, that's the team that needs the focus. So that's how you kinda decide. Right? Yep. That's that's where you make your decisions on what to focus on because we all have very limited bandwidth, and you wanna make the highest impact. Yep. And then as as you make impact through those very high high impact teams, then you can focus on the ones. that are. are. maybe less token. heavy and. And you continue to do, like... So, you know, the way my charter is, I have AI tooling, I have AI enablement, and I have AI transformation, but I also know transformation is the last segment of the whole piece. Enablement is super important, and intelligent enablement is extremely important. Right? So you have to understand your teams, and you have to understand where enablement needs to sit first and then... But you have to have general enablement too. So the. first thing I did after coming to Marvell Technology, I launched an AI training. It's a self serve training on my learning management system. Every employee has to do it. And then that... So these are some of the mechanisms you can use to, to kind of, you know, multiply your own efforts. But you have to do everything in tandem. You have... I have I have all the AI assets. I have newsletters. I have, you know, the internal website. So you can use all these tools to also educate people, which are the nonimpactful cohorts. So you are doing enablement on a larger scale at that point, but you have to have vertical focus. Otherwise, you will never come to the value proposition. Yeah. I wanna take one last question from the group, and I... I'm gonna spin off of a question Manjari said. I believe that's how you pronounce it. And she said, what is the current evaluation or POC cycle for AI compared to times at Uber? And she did say, or are you talking about the recent situation? But I'm more interested in how evaluation and POC cycle for AI has changed over your last three years, like, very involved in this space. Yeah. So POCs, when we started, there were no SLAs. Right? We didn't even know what we were doing in a POC. We didn't even know what success meant in a POC. what are you? measuring? What. are you measuring? What are you looking at? It's fun. It's a great tool, but that's all I know. Right? So I did this very early on, and this is about two years ago back at Uber. Right? What did a POC mean? How do I enable a business domain with a POC? And how do I iterate fast? So we came up with SLAs. The first thing was... I know my manager at the time used to make a lot of fun, saying, Manjari, you have this gargantuan diagram. He used to call it a monster diagram, and it was about business flows. Right? It was about what comes first, what comes next, what comes next, and and and that was my approval process, let alone my POC process. So my approval process at that time was four to six months. And, and and I think that's the problem with every organization. And I was like, in the world of AI, if I'm taking four to six months to even evaluate a POC, It's outdated. I... Too far too behind. Yep. The tool may not exist after six. months. Yeah. So we had to do something about it. And, and that's when I talk about that Stripe team. Right? We came. to an understanding. We downsized our cycle time on everything. And believe it or not, my SLA was two weeks for onboarding a POC. That is crazy. That's that's crazy. But we did... You know, crazy things can be done. Is it... Yep. that's how that's how you become effective. Right? So POCs, two weeks. New onboarding, four to six weeks. That's how. fast we. Get. And and when you evaluate a POC, you always do, like, two to three tools together. Right? It's never one tool. You want to have your use case absolutely defined. What are you doing the POC for? Enterprise POCs are different. Domain based POCs... So by the end of last year, I think, we had 41 AI vertical tools and six enterprise tools at Uber. And this was all an outcome of all the evaluations and all the POCs, and we were running. So at any given point in time, we were running a few or, like, 40, 50 POCs was gonna. say more than a few. Yeah. Yeah. I was just... And this is a random number. Don't quote me on it. I'm just... Yeah. This. this number is not a... Is is not an absolute number that we vetted, but, I mean, I'm just making a general assumption was at this easily easily because we were parallel patting so many, But when, you have a process of, like, exactly what you're measuring. for, exactly what two weeks should happen, it becomes much easier before, you know, testing everything, understand. exactly what those two weeks define. Exactly. Right. And so at Marvell Technology, my first question on my... They already had an intake process set up, and I'm like, hey. What are ISLAs? They're like, we can't define SLAs, for intake. I said, we absolutely can. You have to. we're... Yeah. We have to. So now we're pushing for that, and we are slowly, you know, making our headway there. And it's gonna happen sooner or later. But that's what... That's how you do it. I, love that. actually, I did my first, we are running our first few POCs now after I joined, and I have reduced the cycle time because I did it in a different way. I had all my key partners who had to... Who had questions, valid questions, like security, legal. All of them, join my kickoff call. So you can ask your questions there, and therefore, approvals don't take very long. I got my approvals in the first shot. Right? I'm not going to wait four months to get an approval. So so that's how you kinda expedite. Yep. Okay. I love that. I know we have about five minutes left, and I wanna I wanna end on a question that people can tactically take back. So for women who wanna play a larger role in shaping the future of work... AI at work, what opportunities should they be looking for, and what's one concrete step you would tell them to take in the next thirty days? See, I always, I always believe you do the best you, you know, and always try to do that. Don't move away from who you are just because the world expects you to be a certain way or the environment expects you to be a certain way. Each one of you on this call today brings something really special. That's why you're sitting here wanting to learn, wanting to understand, and so you already have it in you. Just be just identify that within yourself of what that means to you. Right? You identify... Sit with yourself and identify what success means to you. And I have had to ask myself that question over and over again in the last few years of my journey because it is difficult. It is difficult to identify that when there are so many external pressures, when you have pressure from your leadership, when you have pressure from your managers, when you have different difficult situations you're facing. You may be facing situations at home. You may be facing situations outside of work. Right? Different different scenarios. How do you make sense of it and still keep going at what you want to... You wanna do your best, and you want to not measure yourself against others. Measure yourself against who you were and where you want to be. Very easy for me to say, very difficult to do. But if I were to give you one piece of advice, that is what it's gonna be. In the next thirty days, identify that about yourself. And a lot of the questions that are in your mind today, which we couldn't talk about, on the call, will get answered. I promise you that. It's difficult. And I have these conversations with my colleagues. I have these conversations with my, you know, with my with my husband. I have them with my friends. You wanna know what you can drive best at, and you will figure that out out. If you're a good organize... If you're very, very... If organizing is your strength, then you're the best at driving a success in a structured way. Right? If problem solving is your strength, then you're the best at taking a complex situation and simplifying it. And that is what is required in today's environment the most. If I were to tell you what are the two qualities, it is problem solving and being organized. It's very simple. When people ask me... And Jessica asked me this, right, when we were talking about this, casually prior to this call, I always say I'm a very simple person, and I always keep it in the basics. And and I think all of you have it in you. I I always believe that we all have special powers. You gotta identify. If you if you watch... You know, when I watch these movies with my children, you know, all these, cartoon movies which come with a message, if you look at any one of them, what is the message? You be the best you. And that is what we need to keep reminding ourselves. So that's what I would say. You know? There is no concrete. The concrete is defined by you, but you identify that. I love that. That is a wonderful message to end on, Manjari. Thank you. you so, so much for your time today. Thank you everyone who joined. I really appreciate you and your voice and your leadership in this space, and I'm so excited that through this process, I've gotten to know you more. Can't wait to stay in touch and and continue to learn in the future. Manjari, I hope you have a wonderful Friday and an even better weekend. Enjoy your you. celebration, and, we'll chat very soon. Thank you, and I really appreciate you having me. And thanks all. Really appreciate having you all here. Take care. Thank you. Bye bye.