What Does It Actually Take to Land an AI PM Role at a Company Like Meta?

Watch the Episode

Transcript

Shobhit: Ready to move to the next level in your product career? I'm Shobhit from Intentional Product Manager. Join me as we discuss ways to help you stand out in your job search and your career so you can have more impact and make more money. Nikhil, thank you so much for being here as a guest speaker on this video.

Nikhil: Yeah, I'm happy to be here, thank you for having me.

Shobhit: Awesome. So as we dig into AI/ML product management, something that a lot of people are very keen to move into and learn about, take us through your journey — both in your career and then getting into this field.

Nikhil: Yeah, sure, I can start with a brief intro about myself. So I'm Nikhil, I've been in AI product management for over five years now. I'm currently an AI product manager at Meta, where I work on AI for ads. Prior to that I was an AI product manager at Adobe, both on consumer and enterprise products. In my roles at Meta and Adobe I've built and scaled multiple AI products serving billions of individuals and millions of businesses. In addition to that, I'm also a part-time AI educator and mentor — I run a podcast called The Art and Science of AI, where I share my passion and learnings around AI with people who are curious. Prior to AI product management, I also used to advise Fortune 500 companies on big data strategy and implementation — the early days of AI — as a consultant at Deloitte, that's where I started my career. I live in the San Francisco Bay Area, and outside of work and AI, some of my hobbies include reading science fiction and fantasy, and weightlifting.

Shobhit: Amazing — really great to hear about your journey into the space. It seems that you had an AI and big data background prior to your first role as an AI product manager — is that right?

Nikhil: Yeah, that's right — I've been involved with AI for over ten years now. I wrote my first master's thesis back in 2014 on comparing two different approaches to language understanding — there was the rule-based systems approach, which was called formal semantics, and then there's the machine learning approach, which was called natural language processing. And then, like I said, I got into management consulting and worked on strategy and analytics projects there, a lot of big data strategy. I think where I really made the transition to AI and product management was when I went to business school — I went to business school at Yale from 2017 to 2019, and at that time, in addition to the business coursework that we were required to take, I took a lot of elective coursework in computer science and machine learning, and I was learning about topics like language modeling and natural language processing. Coincidentally, this was the time when OpenAI had just launched GPT-2, and I was totally blown away by the possibilities and implications for business and society, and the impact that AI was going to have. That was kind of a transformative moment for me, when I realized that I wanted to pivot my career to focus more in this direction, because I thought this is the thing that's going to be really impacting the world in the next decade or two.

Shobhit: Awesome, so at that point you just realized this — curious what was the journey from that realization to actually getting your first AI product management role? What did you do, how did you prepare — a lot of people are interested in that aspect of it.

Nikhil: Yeah, so, like I said, I was at business school, I was doing an MBA program, and MBA programs are typically set up to help people make career transitions — most of my peer group, including myself, is there trying to transition from one thing to another, people are trying to transition functions, industries, geographies, different kinds of things. So it's an environment that is conducive to helping people figure out how to navigate this journey. I had a lot of resources to help me with this, that's one thing I want to call out — the MBA program really helped. And like I said, in addition to that, I was also taking some coursework in machine learning and computer science, so that's something that certainly helped me a lot — I was tinkering with projects on my own, just learning how to build simple applications, how to build machine learning models and things like that, and that really was something that helped me stand out. I was able to bring my passion and enthusiasm for machine learning and AI to the interviews when I started talking to companies, and that's something that helped me stand out. But yeah, the reality of it is that I did not have any experience in product management at that time — my background was in management consulting, and so I had some domain expertise in AI, I was trying to leverage that and transition to product management at tech companies. So I would say the things that helped me were, one, having resources and a peer group that helped me navigate this transition through the MBA program; second is my domain expertise and passion for machine learning and AI, which helped me stand out; and maybe the third thing is I was very fortunate in that — so the first job I got out of business school was as a product manager at Adobe, on a product called Adobe Experience Platform, which is a customer data platform that's used by enterprise marketing teams, and within that we were building some machine learning and data science capabilities. Adobe was helpful in that they have MBA recruiting programs, and the goal of that is often — they're not trying to find the best candidate for a given role, but they're trying to get talent that they think will grow with the company and possibly stay in multiple roles — so that's another thing that helped me. And then of course there's the standard interview process stuff you have to go through — you have to develop a pipeline of applications, be ready to face tons of rejection, apply to hundreds of roles, get hundreds of rejections, just keep pushing through that. And you have to be willing to proactively go out and network with people, constantly having conversations with new people, trying to learn more about different roles and companies and what's available. And then there's the whole interview preparation process — you have to practice and do mock interviews and make sure you're able to present your knowledge and skills in a way that comes across as having the right kind of expertise.

Shobhit: Just thinking of when you were applying for those product roles — did you focus it on, like, AI/ML product management roles specifically? I mean at that point they were probably fewer than the number that exists right now, given the interest — but was it very focused, or were you open to any sort of product management roles?

Nikhil: Oh, at that time I just wanted any role related to AI or machine learning — I wasn't focused on product management per se, so I applied to a bunch of different roles, I applied to many data scientist roles, I applied to analytics roles. So yeah, my lens at that time was less on product management, it was more on AI, because I was like, hey, that's what I want to do. It just turns out that given my background and experience, like, now I realize product management is a really good fit for me — I don't know, maybe I could have also been a good data scientist or machine learning engineer, but it was harder at that time to convince anyone that they should take a chance on me for that. Product management had a much more clear story, and yeah, since then I've been doing this for five years, I think I found the right fit — I love product management, it enables me to have the level of technical exposure to AI that I want to have, but also it enables me to not focus very narrowly on the technology, but more on what are the problems we're solving with this, what's the longer-term strategy for developing products with this technology, and things like that. So yes, I found a good fit in this function of product management, but it happened by accident — at the time I didn't really know what product management was, I just wanted to work on cool things with AI.

Shobhit: Love it. Okay, so I understand all that — well, so now I'm curious, to whatever extent you can share, you know, assume I know there's confidential campaigns for which you won't share, but like, what are the kinds of things you worked on that have this intersection of product management and AI, both at Adobe and Meta?

Nikhil: Yeah sure, I'd be happy to talk about what I've done so far at these companies. So at Adobe I had two roles — the first role I had was as a product manager on Adobe Experience Platform, and as I said, what the product is, it's a customer data platform that's used by enterprise marketing teams to unify all of their customer data in one place, and then use that to create personalized experiences using machine learning and data science. So for example, a company — say a hypothetical example, say you're a company like Best Buy — you have customer data from many different sources, there's customers who purchase things in-store, or customers that purchase things online, those are two different data sources, you have data about customers who interacted with your website, bestbuy.com, and that gets captured through analytics data — Adobe actually has an analytics product called Adobe Analytics, and people who use that, this was a good fit for them. Then you also have data like, there's call center data, customers might call you and ask some questions or something. So the idea is that companies like Best Buy — but you can generalize that example — have a lot of customer data from different sources, and now they want to leverage all of this data together to create personalized experiences. So for example, next time you call Best Buy, instead of just asking you, hey, what are you calling about, can they leverage all of this data to predict what you may be calling about and give you a more personalized experience, or when you go onto the storefront, on the website, instead of just showing you something generic, maybe they can show you things that are relevant to you. So that was the first product I worked on. Then after that — Adobe has multiple business units, their biggest business unit is called Creative Cloud, where they create software for creative professionals, like video editing, photos, audio, and so on. The other business unit they have is called Document Cloud, where they create productivity software to work with PDFs and documents, and the third business unit is called the Digital Experience, or the Marketing Cloud — so that role that I described, and that product, was in the Marketing Cloud. And then I moved to a new role at Adobe, in Adobe Acrobat AI. At the time, the problem we were trying to solve is that reading PDFs on your mobile device sucks, because PDFs are optimized for large screens, or like A4-size paper or letter-size paper — when you read a PDF on your mobile, you constantly have to pinch and zoom to read it in the right way. So we were using AI to understand the content and structure of documents, and then create a mobile-responsive PDF reading experience, it was called Liquid Mode, and it still is, and I think it's awesome — it's one of the best ways to read a PDF on your mobile device. And now, of course, Adobe Acrobat AI has a lot more than that — with generative AI they're also trying to get into helping people understand documents, like ask questions and answers, and things like that. So that's one thing I worked on. Now, in my current role at Meta, I work in AI for ads, specifically focusing on improving the ads personalization system and the algorithms that determine which ads to show to which users. So our goal is to make meaningful connections between advertisers and users — we want to help advertisers find the right customers who are interested in their products, we want to help users find the right products that they're interested in, and this is a problem space that comes with tons of data, because at any given time there are billions of users, there's millions of ads in the system, and how do you figure out this matching problem — what kind of data can you leverage for that, what are the kind of algorithms you want to use for that. And yeah, that's kind of the area I focus on, and this is also interesting because it intersects with not only machine learning, but there's also a lot of legal, regulatory, privacy concerns that you have to pay attention to, so it's by nature a very cross-functional role, and there are many considerations that you have to balance.

Shobhit: Awesome, great to hear about your work. Now I'm going to put you through just, like, sort of a last set of questions — but given that there's a lot of product managers who are, let's say, somebody with some experience in product management, but they're really trying to figure out how do they get into AI product management, to some extent how do they future-proof their career, like that's the term often people use — and thinking about the impact AI is going to have on product management, product development in general, what sort of advice would you have for them? Would you suggest any trainings? Would love to hear that from you.

Nikhil: Yeah, so I think there are multiple ways in which you can think about engaging with AI — one is at a personal level, and one is at the level of the product, I would separate out these two things. So we're talking about a target audience of product managers, right — so regardless of what your product is, there are certain activities that you do on a day-to-day basis as a product manager, you're trying to do some market research, industry analysis, you're trying to develop a strategy for your product and a roadmap, and trying to come up with requirements, there's a lot of communication and presentation. So all of these things, I think it's definitely helpful to start thinking about how you can use AI to help you be better at those things, more productive, more efficient, so that you can focus more on high-value work — and this I think is general to almost any profession, product management is just an example of this. So I definitely think, whatever profession you're in, please start using AI and see where it can add value, what are the things that it can make you better at, where are the things that can help you scale up, where are the things that can help you automate. So that's one thing, and that's advice I would give to anyone, regardless of the profession. So yeah, within product management, I would look for things that you think are repetitive things that you're doing frequently, things that you could potentially automate or augment with AI, and for me, often it's things like doing research — I need to find some information across a whole bunch of documents, and how do I find the right sources of information. I'm very lucky in that I work at Meta, and we have a bunch of pretty sophisticated internal AI tools to help people be more productive and solve some of these challenges. Of course, I realize that most companies probably don't have specific internal tools for that, so in that case you would have to leverage existing systems like ChatGPT or Gemini or whatever — of course, pay attention to privacy policies and what data you're allowed to share or not, that's a huge thing, make sure that whatever you're doing with your company's data, that it's appropriate, that you're — if you're going to go put that on ChatGPT, make sure you're actually allowed to do that, if you're not, then try to work with some example or dummy data and use the insights from that. I think this problem will be solved pretty soon, most companies are probably going to either have their own in-house solutions, or make deals like ChatGPT, OpenAI, and these companies have enterprise solutions that have all the privacy and data security built in, so pretty soon, I'm sure, most companies are going to have that built in. So yeah, start making use of those, do it for your own personal life as well, in that case then you're not restricted by your company's data policy. A simple example that I use — as, you're a content creator, you create YouTube videos and things like that, and you know it's a lot of work, every time you create a video you have to edit it, and then you have to extract a title, chapters, descriptions, show notes, and things like that. So for myself, I automated that, I created a custom GPT where it has knowledge about my podcast, and then I just upload a new podcast transcript to it every time, and it automatically gives me all the info I want in the format that I wanted — it'll suggest here are the episode titles you can use, here's some descriptions you can use, here are the chapters, and so on. So anything like that, anything you're doing repetitively, think about how to automate that, and I would say, both in your personal and professional life. And then there's the other dimension of, okay, as a product manager, sure, I'm using AI for my personal productivity, but you're also now thinking, maybe I want to think about transitioning to working on an AI product, or incorporating AI into my own product. And so, yeah, that then is different, that's more of a career transition type thing, especially if you want to change your role to work on a different product that's more AI-focused. I don't know if that is — I do see a lot of people these days want to make that transition, but I'm not sure if it's super necessary, because I think pretty soon all products are going to have AI incorporated into it in some way, so you could also just think about, well, how am I incorporating AI into my current product? So that's something you do, but yeah, maybe there are other arguments you could have, like, you could say, okay, for my personal growth and learning, I want to go work at a company that has a more established understanding of how to use AI, and so I want to work there and then bring the learnings back. Yeah, I mean, I think that's — it's similar to reasons why I think PMs often want to work at big tech companies, to help accelerate their career, and then you can kind of go to a company in any other domain and then transfer your knowledge of best practices and things like that to the other domain you're working in. And so, yeah, I think if that is your goal, it does make sense to try and get some experience working on AI products that are being developed by some of the more sophisticated AI companies. And I guess if your question is, well, how do you make that happen — I think a lot of it is just general PM career guidance will apply here. One thing I always advise people is to think of your career transition as a multi-step journey, and it's often not a single step — like, often people who are not product managers right now, they're doing something else, they might want to know, well, how can I become an AI product manager at Meta, and I think one of the things to recognize is that that may not be something you can do directly in one step. Like, think about, if I were to break this down into a multi-step journey, what would that look like — so maybe, let's say, for example, you are a product designer at a fintech company, think about moves you can make that minimize degrees of transition. So there are many dimensions to your role, right — there's the function you're in, which is like product management, product design, consulting, whatever, there's the industry you're in, there's the type of product you're working on, there's the technical domain you're working in — so the easiest transitions to make are ones where you are minimizing the number of things that you're trying to change. So if you're like, hey, I'm keeping everything fixed and I'm just changing this one thing I work in, you know, I work in B2B SaaS, I want to go work in B2C SaaS, doing everything else the same, same industry, same manager, so that's the easiest. The hardest transition is where you're like, okay, I want to change a bunch of things, I want to change the type of product I'm working on, I want to change my function from product design to product management, I want to change the type of company I'm working at, from a startup to a big tech company. So I think, yeah, you should balance that — if you notice that the transitions you're trying to make are like many, try to break that down into smaller chunks, and see what are more reasonable transitions. And specifically, I think for the transition of, I have no experience in AI and I want to work on a more AI-focused role, for that part of the transition, try to find ways to show your understanding of this domain, and it could be as simple as the kind of things I mentioned earlier, which is, find ways to incorporate this into your personal and your workflows, and maybe create some automations, go publish them on your GitHub profile, or there are many no-code platforms now where you can do that, and that's a really simple way of being able to demonstrate your passion and expertise in this area, and show that you're actually able to use this stuff to solve real problems.

Shobhit: Yeah, become a power user first, before you go and try to understand the tech.

Nikhil: Yeah, yeah, I think so.

Shobhit: Awesome, Nikhil, thank you so much for sharing your journey, sharing your own transition, and then also advice for others. I totally resonate with that example of automating YouTube stuff — I follow the exact same thing, and it's so helpful, and it's taken so much repetitive work off my plate.

Nikhil: Awesome, well, I hope whoever is listening to this finds this helpful. Oh, and also, if I may, I'd love to make a last plug for my podcast, The Art and Science of AI, where we talk about various topics in understanding the science of how AI works, and the art of using AI to kind of reimagine your life or your business, and we have a lot of examples, we talk about how to build automations and workflows for yourself. So yeah, if you found this helpful, please check that out.

Shobhit: And I mean, just learning from that would be a very logical first step for folks who are looking to transition into their product — thanks again, Nikhil, appreciate it.

Nikhil: Yeah, thanks for talking to me.

Shobhit: Hey, be sure to check out our website at intentionalproductmanager.com to see how you can level up in your career.