This Mahesh Yadav podcast transcript captures a comprehensive conversation about building AI agents and breaking into FAANG companies as an AI product manager.
Check out the conversation on Apple, Spotify and YouTube.
Introduction and What Makes an AI Agent PM
Aakash: Mahesh, what are we going to cover today?
Mahesh: Today, I’m going to show you how to build agents—not only a fancy front-end in Lovable, but also a back-end that you can scale with multi-agents. Also, I’ll talk about how to get $300K+ jobs in AI, especially agentic AI, and share any secrets that I’ve learned throughout the years.
Aakash: What does it take to become an AI agent PM?
Mahesh: I know it sounds hard, but it’s three simple things. One, we’re looking for people who have built in AI. Second, we’re looking for people who have done AI PM transactions, which includes how you dealt with data, how you dealt with models, evaluations, and scaling by iteration. And last is just generic skills around scale, ambiguity, and experimentation. Because this is new, and these three are becoming essential or table stakes now.
Building Your First AI Agent: A Live Demo
Aakash: So I don’t want people to have to wait too long. Let’s get right into how to build these AI agents, both backend and frontend.
Mahesh: Sounds like a plan. I’ll be using two major tools. We do this with new people to show them how they can build their AI agents.
For the backend, I’ll be using this tool called Langflow. Langflow is a tool which allows you to build no-code, highly complex or simple agents. So let’s start building our backend first. I’ll be building a simple tool which you can use for competitive analysis.
As a PM or as anybody who’s trying to build or analyze markets, you need to analyze competition. So we’ll be building a tool which allows you to do competitive analysis.
The Architecture: Input-Output-Process
Mahesh: Let me tell you how we’re going to build it. We’re going to use Langflow, which is a no-code tool. It allows you to build highly complex or simple agents without writing a single line of code. And for the front-end, we’ll be using V0. This is like Lovable or Bolt. You just write prompts and it creates an awesome web app or front-end for you.
Then I’ll show you how you can combine a very powerful backend in Langflow to a frontend in V0.
Aakash: Before we even dive into that, how did you choose Langflow and V0?
Mahesh: When I was starting, I was looking for no-code tools because once you start writing code, you need to learn editors, you need to learn coding skills. And suddenly the barrier to entry is so large that most people don’t cross that barrier. That’s why I was looking for tools which are easy. So one is ease.
Second is capability. Langflow was quite capable. It had integration with all model providers. So I can choose Azure, AWS, or Gemini, Google models. And third is this ability to get access to the code if we want to go to production.
Setting Up the Agent Components
Mahesh: If you’re building your agent, the first step you need is: which competitors you’re going to compare. So you need an input as the name of your competitors. What tools your agent will be using—it needs at least a search tool so that it can do all the search on the internet and find the information you’re looking for.
And then obviously you need the agent to process, which requires you to have a prompt with special instructions. I’ll show you tips and tricks on writing these prompts. And then what is the output? So input is name of competitors, output hopefully would be a table format, and the tool is a search tool.
The Art of Prompt Engineering
Aakash: The system prompts in particular for your AI agent are everything.
Mahesh: Let me talk about what it does and how you can write a good prompt structure. The first thing you talk to your agent about is your role. So you’re saying, “Hey, you’re a professional business analyst with expertise in corporate research and competitive benchmarking.”
Now your instructions: “Compare the two companies given—Company A, Company B.” These brackets are telling it that this information is coming from outside. Once you put it in brackets, the prompt structure will change and it will open up two ports so that you can pass this information from outside.
Then you specify exactly what you need in your output. These models don’t know how your formats are, what your company needs, what your report needs. So give them very detailed instructions.
And then the guardrails. If I’m talking to you, or if I’m interviewing you, or if I’m seeing how well you know—if you have a good structure in your prompt like role, instructions, guardrails, or even tool sections specifically—I know that you have learned AI well.
Connecting Backend to Frontend
Mahesh: You can make it as complex as possible. This allows you to do RAG, multi-agents with CrewAI. We have labs where we’ve gone ahead and even put agentic RAG with graphs or did some fine-tuning also.
But today, let me show you how you can take this and connect a Lovable or V0-like agent with Langflow so that you can do all the fancy work on the front-end, but your back-end is where your AI is and that can remain separate.
API Integration and Testing
Mahesh: If you click publish, you click API access. They allow you to access this as an API, the whole flow. To build this end-to-end, you’ll need the response also. For that, you can use this tool called Postman. Postman is a tool which you can use to test back-end APIs.
Once you know what command to send and what the response is, you can go to a tool like V0. Then I’m going to give you a prompt: “Build me a modern responsive landing page for a competitor comparison tool using provided APIs. It takes two inputs, Company A and B, and a compare button. It calls APIs, resolves issues like CORS…”
In the past, a developer needed to read maybe 20 API documents to get here. But now all you’re doing is copy-pasting the response. And that is it. You can send this request right now. And hopefully in two minutes, your front-end will be ready and you’ll be ready to publish your app.
What Makes This an AI Agent vs Regular AI
Aakash: So what makes what we just built an AI agent versus a regular AI product?
Mahesh: What makes it an AI agent is three things. One is that it uses tools. Second, if you look at the instructions, it is taking a role and a goal. And unless that goal is achieved, it will keep on trying things and it won’t fail. And third, which we have not done, is connecting knowledge.
And just—I said three, but one more is the guardrails. It can put its checks. So it can assign things and it can have memory. So it can do many interactions, learn from each iteration, and then keep on improving until you achieve your goal.
That’s an agentic capability, which is not an AI capability. AI services are more like: you give me speech and I give you text back. It is only single purpose. Here it’s doing competitive analysis and you can keep sending different companies, different verticals, and it will continue to do that thing. And it can fail, but it will recover and you need not worry about it.
The Evolution of Agentic AI
Mahesh: Let me give you a short history of agentic AI. I was lucky enough to start working in agentic AI for a while, and this is the first time everyone can remember that moment when you first heard about ChatGPT. It was such a compelling proposition that you took your data to the product.
That has never happened in AI before. I was trying to sell AI-based products seven or six years before ChatGPT, and we had to sign hundreds of NDAs, go through corporate loops to get people’s data. But this was one product where people actually took their data and copy-pasted it in ChatGPT and said, “Analyze it for me.”
So that was the moment where we started seeing success in chatbots. This is where we solved the Q&A problem.
The Three Waves of AI Evolution
Mahesh: In 2023, what happened is we said, “Okay, you need not take your data to ChatGPT.” The big giants woke up like Amazon, Google, or Microsoft, and they launched some version of Copilot in their products. So now you can have the product, you can write, swipe and ask questions or do things which were not possible before.
And then 2024, we reintroduced agents. I was lucky enough to work on the first framework to build agents at AWS. And there we said, “Okay, Q&A is cool, but what about I let you connect tools or make changes in the world or change the state of the world by calling APIs, doing search, writing code—all these tools, if you connect to these Q&A machines, can they start doing things?” And that’s agents.
And then in 2025, we’re making them multi-agent and multi-modal. What does that mean? That it not only takes text, but can take images, which I was showing you on V0—how I can just paste a picture and say, “Hey, make it look like Gartner’s website.” Or I can send my input through audio also.
Also, we’re saying not a single agent is enough. Maybe connect many agents to do complex things. One good example of multi-agent is the AI coding agents. This is not a single agent. One agent is thinking, the other agent is writing the code, another agent is testing, and all that combined are able to do complex things.
So we have gone from simple chatbots, which were amazing—can summarize, can generate poems and all—to fully blown multi-agents which can write code like all of us do, create web apps, or do PhD-level research for us. That’s our journey so far.
Mastering the Vibe Coding Interview
Aakash: Now let’s talk about these vibe coding interviews, these AI coding interviews. We just demoed an end-to-end workflow for people. A lot of people are going to need to execute these in interviews because places like Google are bringing on these interviews. You worked at Google. You helped them write some of their first AI guidelines. What do FAANG companies like Google look for in these interviews?
Mahesh: AI is new, right? So what we’re looking for are people who have actually built in AI. Because what we want to avoid is people who are setting the expectation too high or setting them too low. And we don’t really care—we understand that you have not worked in production. But have you actually built an end-to-end workflow and maybe 10 people tried it? How many iterations of that?
So we’re looking for builders and not people who know just frameworks and emotion. That’s the most exciting part for people like me.
The Three Key Skills Companies Look For
Mahesh: Second thing we’re looking for: people who have done interactions in the AI-specific world. Let me explain. What that world looks for is: Do you have a sense of how to interact or get your data or handle data responsibly? Can you understand how these models are designed and which models are good at what? Can you evaluate an agent or AI, or can you work in iterations?
And third, how much have you achieved at scale? Not in AI—maybe with some other example, you can bring your cloud scale or if you pivoted in mobile. But we’re looking for people who have seen one technology transition going from one change, which is big enough, and how they handled it.
If you have two of these, I think you have a really good chance to get into any of the FAANG companies because they are desperate, to be honest.
How to Ace the Vibe Coding Interview
Aakash: So this PM interview type that has popped up, this vibe coding interview—they often ask you to vibe code live in the interview. What’s the right way to handle this interview type? What’s the right framework and structure to ace it?
Mahesh: When we’re asking you to vibe code or when anybody asks you to do vibe coding, the requirements have not changed. We’re not looking for your technical skills of how good you write code or do you understand what is written. What we’re looking for is: what are you adding in your code?
It’s just an easy way for us to create something together and help me judge how you think as a product manager. So when you’re vibe coding, do not forget your product principles. That’s my lesson number one.
Second thing, please play with these tools. They’re not very hard. So what I’m looking for is like the prompt I showed. Do you structure your prompt? And that tells me that, yes, you have done some interactions with AI. And second, you understand the limitations of these models—that they can’t take jumbo instructions correctly.
And third is: more than you just show me how cool this thing is, you show me these iterations. That tells me what is your taste. So I’m looking for two things, to be honest. I’m looking for taste. And can you improve with iterations? Can you evaluate and then improve it?
So show me maybe three things: one, show me how you think and show me in your prompt how you think. Second, once you build it, show me what kind of user insights you’re bringing in your second iteration. And hopefully third iteration, show me some feedback loops that I will make for this, which shows me that you have a sense of data and AI in you.
I think that could be a very good three-step framework that people are looking for when we ask you to vibe code. So just take it like a canvas to show your PM skills.
For more details on breaking into AI PM roles, check out our complete AI PM job search guide. Mahesh’s insights on vibe coding interviews are particularly valuable for candidates preparing for FAANG companies
Cart Before the Horse Development
Aakash: One of the things that we talked about in our pre-show conversations was this idea of cart before the horse development. I loved your take on this. What is cart before the horse development?
Mahesh: I think now all of the PMs can hate me after this. But the idea is this: PMs, in every room they go to, they tell engineers, “Hey, this is us and we will tell you don’t put cart before the horse.” But in AI, that’s changing because that’s the whole idea of having a PM—because they decide where the horse goes and then build the cart.
But to be honest, I think this is the first time we’re actually in a position to change that idea a little bit. Because what’s happening is three things.
Why Cart-First Development Works in AI
Mahesh: One is the cost of prototyping has gone 100x lower in the last two years itself. So building something as you saw is very cheap and anybody can do that. That’s number one.
Second, the customers really don’t know what to expect out of AI. And third is, there’s this idea that so many possibilities, so many ideas, so many things are possible. And if you’re going to do six months of research to figure out which problem to solve, maybe somebody who did something else will solve it before then. So this FOMO factor.
The New PM Workflow
Mahesh: The old method is: I take this problem. I do research for three months. I write this awesome piece called PRD. In another two months, I get these approvals, and then I hand it over to engineering. Hopefully I’ll show up when business needs me or when it’s done in another six months, and we launch one product or one feature every year and I get promoted. That’s the old world.
The new world is: the PM actually talks to customers for two weeks, three weeks, creates a prototype, shows them what’s possible. And then iterates on the prototype maybe for two or three weeks. Then writes a very small PRD with detailed user experience—this is what happens on this click, this is what we should show—a clear user experience with ideally some prompts. And then evaluations, which is how we’re going to evaluate what the quality expectation is.
With these three things, they can hand it over to engineering. Engineer makes it possible and makes it production-ready. And then you iterate every three months, and in a year, you have built a product and actually made it make money and made it what exactly you could never think or nobody would ever tell if you hadn’t done that.
And that’s the new way of doing AI PMing. And in this case, you’re building the cart first and then working with the customers to figure out where the horses will go. That’s the idea of cart-first AI product development.
Breaking Into FAANG: Mahesh’s Personal Journey
Aakash: I want to shift into how you broke into FAANG because a lot of people want to break into these companies. They want to get this $300,000+ PM AI agent’s job. So let’s start with your own story. How did you break into AI at Microsoft? I believe this was back in 2016.
Mahesh: First is having an intuition for a new technology. I had that. I was working in Azure IoT, and I thought maybe this idea that you can, instead of hard coding or doing logic-based programming, this whole ML thing would be a cool thing.
But as you may all be feeling, there’s not easy entry. There are barriers of terms like gradient descent, local minima, math—you need to be this research scientist. It was even worse in 2016.
The Weekend Hustle Strategy
Mahesh: For me, I just wanted it so desperately that I started hanging out around Microsoft Research campus. I offered my services on Friday, Saturday for free, and then I came up with one idea of building a Vision AI dev kit, which was a camera that can inference on edge but train in cloud.
I offered that to our Azure Machine Learning team. And I was a dev then. And they said, “You need to test.” And in Microsoft, if they tell you to test, it’s not a good deal. But I took anything. I said, “Yeah, I will test it.”
In three months, some people dropped something. I took on that code, then coded. And eventually I became the TL for that. I shipped that product end-to-end. That was my story. I was the guy who actually built Vision AI DevKit and shipped it end-to-end to production. Satya took it at our 2017 or 2018 Build on stage.
The Marathon Analogy
Mahesh: After that, it’s just a story from there. I knew a lot of things like how chip acceleration works. I knew different architectures. I knew how to train them in cloud. I knew how to optimize them to run on small hardware.
After that, it’s just iterations. I think after that it becomes like a marathon where a lot of people run with you, and if you get to do the first good project, you get a good lead. From there you can run the marathon correctly without running into people or getting stopped every five minutes because nobody’s moving in front of you.
So my two tips out of my own experience are: be ruthless and passionate. Take anything that’s coming in ML and AI your way or create your own paths. And once you have them, don’t take them for granted. Work twice or three times harder once you have those opportunities.
I’m pretty sure it will change your life. It has changed mine since I was able to double my salary every other year after that.
The Transition from Engineer to PM
Aakash: Talk us through how you made it from dev to PM.
Mahesh: When I was doing this Vision AI, I also realized another insight, which was: in AI, I believe that building things will get easier and easier. But what to build and how to work closely with customers to build the right thing would be a skill that would save a lot of time for organizations.
To be honest, I hated what PMs did because they could just tell me and then I had to spend six months and I couldn’t question a lot. But I found them little vague, so I never wanted to get into PMs.
But the stakes were so high for me that I did two iterations and things didn’t work out. And I don’t want to blame anyone. So I said, “Okay, let me figure out.” And my God, it’s so hard to figure out what’s the right thing to build. It’s super easy when somebody tells you what to build. But figuring out what is the right thing to build, especially for a large company when the stakes are higher and you’re at L7 levels—you’re playing with at least 300 to 400 people’s lives and you’re almost playing with $50 to $60 million worth of decisions you are making.
Why the Manager Wanted Him
Mahesh: The PM manager actually said, “Mahesh, I want you in my team. Can you join my team?” And I said, “No, I can’t join your team because my visa has this problem…”
But he kept pushing, insisting on me. And why? To answer your question: because I was super passionate about customers. I was working backward from there. I was there to solve the problems. I could prototype real quickly. And I cared about business.
I cared about making money for the company. I cared about what will matter when all AI dust settles. And I cared about these three things, which is: Are we building the right thing for our customers? Is it going to make money for us? And how can we bring people together to build it?
I think these three skills were essential for being a successful PM. And if you show that to people, if you do that without giving that title—it does require time, effort, and energy, to be honest, because I needed to still code. But I went ahead and did that anyway, because I wanted to code what will matter, not code what things won’t matter.
Building Credibility Through Customer Insights
Mahesh: If you’re struggling, start doing these three things, which means writing a lot about what is the problem. Just don’t write PRDs because your PM friends will get offended. But write your customer insights.
I had this customer insight newsletter I started. Because I was debugging each customer problem with Vision AI DevKit, I would write as if I talked to them: this is their problem, this is the model they’re using, this is why this fails, this is what our system has gaps, this is what AI as a whole has gaps. And that’s why I just patched it right now. But this could be the next thing for us.
One example of that is ONNX (Open Neural Network Exchange). When I went to Azure Stack, which is this idea that you can run AI on-premise rather than sending everything to cloud, we were dealing with government contracts. We took Intel FPGA chips for that design, not NVIDIA.
Then I learned quickly that these architectures—we were doing vision models, so MobileNet, YOLO, and each time there was a new architecture that comes, Intel can’t catch up and it works out of the box in NVIDIA. So I wrote about these and then we came up with this idea of ONNX.
The idea is that Microsoft built a layer which allowed any architecture to automatically do accelerated computing across any chips. So you can train a model in TensorFlow or PyTorch. You can convert it into ONNX that Microsoft provides you with an open-source framework. And then this model can run on any hardware accelerated.
Once you start making these kinds of contributions, nobody will stop you. Either you’re going to become a PM or you’re going to become a GM through the engineering ladder because people want people who are taking broader impact and broader challenges than what is given to them.
Cracking Multiple FAANG Interviews
Aakash: And then after that, you worked at so many of the other FAANG places—Meta, Amazon, Google. How are you cracking all of these FAANG interviews in a row?
Mahesh: Once it was very hard, to be honest. I was born in Microsoft. I was supposed to die in Microsoft. Microsoft is such a family, especially on this side of Lake Washington. If you live in Redmond, you live in a Microsoft bubble. You go to restaurants with all Microsoft people, you go to the gym with Microsoft people, all your friends and all your kids’ friends are Microsoft. And it’s just hard to leave that place.
So first time when Facebook reached out to me, at that time I was working on inference, but I wanted to go into training and they had an awesome opportunity for me. So that time it took a lot of time.
What Made the Difference
Mahesh: But after that, I think AI has done a great deal for me. I had this different skill set which was in demand across these four or five years where I’ve done these transitions. So I would say that played one role.
Second role is that I was able to just articulate well my stories, show scale, show how I handle ambiguity, and third, this never-giving-up attitude which I think these companies are hungry for.
Because once you enter these companies, the people there are just too hard to work with and things don’t move fast. And if you can say that I’m going to come and make things go fast, these PM managers are hungry for you, especially in AI because everybody had that FOMO pressure.
And this guy knows what we can build. And he has done iteration outside. And he’s hungry. He’s impatient. These three, I think, just made it so easy for me.
Interview Preparation Strategy
Aakash: And how did you get so good at the PM interview though? Like the PM interview, case interviews, what was your approach to that?
Mahesh: I think I watched all your series, read all your blogs. I truly followed you and others and I was able to keep up with what’s out there. There’s a lot of good content or summary, but I did—I generally don’t stop at LinkedIn. I take what you guys post and then I look at the links and I read them.
I was reading—there’s another guy named Lenny—and he posted about AI jobs and I was trying to figure out, okay, what do the trends look like? Is this going to be the right thing for us, for the economy? And then I read like five more reports.
So I think staying on top of this—and don’t do your own research. Let people like Akash, Lenny work for you so that you can get it from an abstract form and then scale it to your own world. That’s one thing I would suggest.
Second, just having stories where you show these three things, which is scale, ambiguity, and what is your impact on business. And show it, say it succinctly, and let the interviewer run your interview, but land these ideas.
And third is just having an opinionated view on things, I think was helpful for me. Don’t be a pushover if they push you. I’m very respectful and I listen to them, but I try to say why I thought this. And you can go with me three levels down and I’ll be able to tell you that my fundamentals, my model of the world is not very bad. And then you can say, “Okay, maybe in this case he might be right.”
Comparing the FAANG Companies
Aakash: So you worked at all these companies. How would you compare them? How would you compare Microsoft, Amazon, Meta, and Google? Which one is best for which people?
Mahesh: Let me give it a shot. I’ll tell you what I loved about each company.
Microsoft: The Innovation Dreamland
Mahesh: Microsoft is a dreamland for people who want to go build without caring how it will make business. I don’t know. It’s just ingrained in Microsoft to just innovate without figuring out how we’ll make money.
If you want to be a PM, if you want to build a lot and try taking a lot of shots, then go to Microsoft.
Amazon: The Business-First Approach
Mahesh: The opposite of that is Amazon. If you’re working in AWS specifically, you have to be very clear how you’re going to make money. If you want to be a PM who owns profit and loss statements, who actually cares a lot about business and ready to die for that $1 thing and execute at the speed of light, go to Amazon or AWS specifically. AWS AI teams—so much fun there for that one aspect. But they won’t let you innovate a lot. That’s the downside.
Meta: Innovation with Great People
Mahesh: That brings you to Meta. Meta is a company where you’ll go if you want to innovate a lot with the right amount of people. Here, one thing that I still remember is that engineers there are world-class. And they will come to you and say, “Tell me what you want me to build.”
And you say, “Dude, I’m not sure. I have these three ideas and all are like 20% sure that they’ll be successful.” He said, “Can you give me one? And I will build it. And then you can launch it. And then if it doesn’t succeed, I won’t even tell my manager.”
They will take on these experiments. They will build it over the weekend. I will be able to launch it on some sample set. And if it works, all they want from you is create a post and publish it. And then they will keep coming to you.
But if you can’t give them good ideas, if you fail three times back to back, they’re going to the other PM and they never come back to you. And that’s your death in the AI world if you’re working for Meta.
So Meta is a place where you want to have good iterations and working with great people to do the next level thing. Also never worry about money inside Meta. They have a lot—it’s a very rich company. And it’s an awesome company inside. Outside people think badly about whatever they do. But inside, it’s very homely and awesome.
Google: User Experience Masters
Mahesh: And last one, Google. Google is an awesome company for user experience. They just are obsessed with user experience. If you want to learn how to serve anything on user experience, there’s so many experts there—what a copy should look like, where a button should go, what happens from this screen to this screen, and how can we create magic between this experience, even if there’s a five-second delay. They are obsessed with that.
And second is, they give you a lot of time. So it’s like Microsoft, but they care about money and user experience. A little bit of money—less money, not crazy like Amazon. So a little bit of money, but a lot about user experience. So if you want that balance, go to Google.
The Market for AI Agent PM Jobs
Aakash: So we’ve been talking so much about how to break into these FAANG companies. And we started with these agent PM jobs. How many AI agent PM jobs are there really out there?
Mahesh: Let me show you. I got this data from this site called Trueup. They did this analysis and they show you all the data about how these jobs are distributed.
This is the data on jobs in agentic AI or AI specifically. These are all the job postings that are open. Don’t look at the numbers because people who scrape these jobs have some problems, but look at the trend.
You’re looking at a lot of jobs in PM, then it dropped, and these are all the PM jobs. If you go to the next slides, you’re looking at all the AI jobs. And now you see where the growth is coming from. So this is Trueup and you’re looking at 25K AI jobs.
And then this is the last one where Lenny did this analysis, working with Trueup. And here it shows that if there are new jobs, the likelihood of that job in AI is double than the likelihood of that job in PM space.
If you look at this comparison, this is growing two to three times more than what your traditional PM jobs are growing.
Compensation Expectations
Aakash: What is the salary and compensation for these AI agent PM roles?
Mahesh: Easily, you’re looking at Level 6 or Level 7 roles at FAANG companies. You can expect anything from $750K and above. I think that’s a normal pace these days. And if you’re looking for Level 8 and above, you’re looking at $1.2 to $1.5 million. That could be easy.
And I’m saying total comp—your joining bonus plus your stock. And even if the stock remains the same, that’s the comp these days across these companies. And most of the startups are trying to give even better.
So if you’re going to OpenAI, at Level 6, Level 7, you can even expect $900K. Netflix recently posted a job. I think they posted some $790 or $950 across these jobs. And that’s normal for top 10 AI companies, and I think that’s why people are excited about it or people spend so much time and years preparing for that.
Can Anyone Become an AI PM?
Aakash: Insane numbers. Can anyone become an AI PM?
Mahesh: I think so. Don’t expect to become an AI PM if you have never worked in AI before in six months or six weeks, to be honest. It might take you a year to get to this level where you can start interviewing at top companies and get to $700+ salary.
But if you want to get to $300+, I think you can crack something in six months. And 18 months is a really good time to get to FAANG if you’re starting from zero.
I think it’s a great time to do that because there’s a level playing field. There’s a lot of AI PMs needed, and whether you were at Google before or not doesn’t matter, or whether you were at Meta before or not doesn’t matter because we need that talent.
And if you are outside and have been iterating, solving ambiguous problems and building products with iterations and not by research, then I think this talent is what the world is looking for. And if it’s not happening, just hold on to it. By the way, it took me—AWS told me no when Meta took me. Google also said no. So just a secret for you that it doesn’t always work, and then you just get double salaries.
It takes time, but if you’re working on something that you’re passionate about and you believe in the tech stack and you are working as fast as you can afford to, 18 months is a really good time for anybody to just get in and get to the top 1% of AI PMs.
The 18-Month Roadmap to AI PM Success
Aakash: What is that 18-month roadmap to becoming a FAANG AI PM?
Mahesh: So let’s divide it into six milestones. Your first milestone is understanding and building your first good prototype.
Months 1-3: Master the Fundamentals
Mahesh: So what does that involve? That involves: do you understand what goes behind the model, this intelligence piece? So this intelligence piece is your LLMs. You can call this GPT-5 or Claude.
How this piece is built, then how you connect it with your own knowledge of the company, your own databases. So this is the thing that we all have intelligence and this is coming in machines these days. This is how the world works.
Knowledge that is not in the model, which is: what are the locations in your company? What you like, where you have visited in the last year if you’re building a travel agent.
Then your memory or signals. Then your tools, which can change the state of the world. So this is how the world works—current state of the world. And then your tools, which is how can you change the state of the world.
And then guardrails, which is what you should do and what you can’t do. Or putting all the checks and laws. Or even checking by feedback: did we achieve the goal or not.
And then bringing this all together to build a real working agent. That should be your goal in the first three months. That you understand these concepts in and out. And you have built your first agent with all of these ingredients in it.
Months 4-6: Build and Evaluate
Mahesh: After that, what you’re looking for is either starting your own thing on the side, which is a real good problem you have, which can be any problem under the sun. I don’t care. Do you have PhD-level expertise in that thing? Does it have unstructured data and complex decision making?
And if it has all these three, just start solving that problem. That’s your next three months. And you should have at least 10 to 20 free users on that. That will teach you how to evaluate these things.
So the next thing we’re learning is: first three months you learn concepts, then you go and build and evaluate.
Months 7-12: Scale to Production
Mahesh: Once you reach that, that’s your six months. That’s your first iteration. And after that, next six months, I really want you to get it into production. Whether you are hiring your own team of five engineers, maybe somewhere in India or somewhere it doesn’t cost that much. And you are actually making it production and going from 10 customers to 100 customers.
Another idea there is work for a startup for free. People will love to take that. And everybody wants people who have this much done and help them give them free hands. So that’s your next six months: build and give people the right thing. This will teach you how actually this build-and-evaluate iteration will work and bring the ambiguity and scale factor to your world.
Months 13-18: Target Your Dream Companies
Mahesh: Once you have that, you have hit a year. Then you start actually doing a lot of—after this, you’re trying to deconstruct any companies that you want to target, their products, and seeing what feature you will build. And you start working in their open communities.
You run evaluations for them free. By this time, you should have your evaluation frameworks. And you’re running evaluations and telling them what are their gaps. And you’re doing it for top 10 companies that you want to work for, for their top 10 products, which everybody cares about.
Whether it is OpenAI, Anthropic, Meta, Llama—start doing that. I’ve done that and it does wonders. And then you create a small community around it and you’re building them and you’re saying, “Hey, you know what? I got these three engineers. I came up with this feature list and I have checked it into your codebase.”
At that point, Meta won’t hire you, but Google will hire you. Nobody can stop you. Next iteration, you’re even doing better than their PMs.
The Open Source Advantage
Mahesh: And the beautiful part is all these companies have open-sourced their model or some version of their model. Even GPT has an open-source model now. And you can take that and show them how GPT-5 is so bad at instruction following and take the open model and fix it and do a side-by-side eval and publish it as a paper.
And you can do that once you reach this one-year timeframe.
Real Success Stories
Mahesh: After that, our students—people who have done our courses, have been in our community, have done all the paths—they get jobs before this step. But if they don’t get it, I know a guy who just hacked the whole website and showed them their prompt and said, “Hey, maybe you want to do these checks. Otherwise, people will see your prompts.” And he got his first gig like that. And then he worked—now he works for Carfax, building their AI stack.
So it’s not like people are desperate for you, but you need to be desperate for learning. And that’s your 18 months, month on month. And every three months, you should be iterating after this.
Aakash: This was the clearest, most actionable roadmap I have gotten yet. And I’ve asked quite a few people this question. So make sure you guys were taking notes on that.
AI Agents for Regular PMs
Aakash: I want to talk about this final angle of AI agents, which is: if you’re just a regular PM, not an AI PM, what AI agents should you be building to improve your efficiency on the job?
Mahesh: One AI agent which all of us need is the guy who looks at our customer interactions. So we live in a multifaceted or multimodal world and we go to a lot of calls where our sales or our friends or teams talk about this.
So always have a single agent which goes through every channel, every interaction where you can get feedback on your product, processes it and does something. I think that’s a missing agent. I haven’t seen any tool. I’ve seen some tools here and there. I think you should build it and iterate on it. So that’s my number one suggestion.
Three Essential AI Agents for PMs
Mahesh: My number two suggestion is build something that you can run A/B tests at scale, which is this idea that before what you want to run, there’s never enough people to run A/B tests. And to be honest, when we ran them, it was such an insignificant sample to make any judgment calls.
But we live in AI now and you can create thousands of different personas with small variations, put them into groups and show them your product and maybe run quick simulations of A/B testing or just simulation of who is going to use their product most or love the product most.
So one is, this simulation should teach you how your product fails and for what customers. And second thing, the output of that agent is that it tells you which personas will love your product—which, by the way, will be the people who have finished the flow faster, have stopped at the key points, and clicked the feedback.
The Reviewer Agent
Mahesh: And I think all of us should build a reviewer. People who are new in AI lose a lot of their brand in the beginning because they have these documents which are immature. They have used terms loosely, and not to their fault—they are just new.
So I will build a reviewer. I think people who are building PRD tools or create PRD tools, I think they’re really foolish. They don’t understand how much is about credibility of the person. And their intuition goes in that PRD. And if AI generates it and you think that can solve—or that’s how the organization will make decisions—I think we’re not there yet.
So I don’t need that. What I need is: I will write the PRD and I will check it with AI. But I want to have a reviewer who has looked at my manager’s 20 reviews, how he commented, and can automatically comment on my PRD. In his style, what he cares about, what he doesn’t care about.
And if I build that reviewer, everybody should be able to use it. Just upload whoever you are reviewing with, take their previous 10 reviews, upload them, upload your document, click a button, you got the first review. You address things and then you show up.
I think that people will use and will be a great thing. So A/B tool for testing at scale to run simulations of your customers and get insights, which is based on this thing which actually watches different channels and gives you insights. And third is a reviewer tool. I think these are the dream tools which I want to build or use, or I think everybody should build and use.
The Business of Building AI Communities
Aakash: So we talked about how well these roles are paying, yet you’ve chosen not to continue on the path of these roles. I have to ask, this is the hot question now, you’re on the hot seat. How big is the business of Mahesh?
Mahesh: I think I’m very comfortable with the community and things that we build. I won’t give you my exact numbers, but you can look at how much our course sells for. We have grown 40% cohort by cohort. We started with 40 people and we have grown 40% each cohort.
We’re on our eighth cohort and we just served—we are serving 180 or so people. And we have divided it into two because I never wanted to have more than 70-80 students in one group. Because I really wanted to have personal interactions and know them by name. So I run two classes now.
So that’s how we’re making money. But I’m also working on building my own thing. And the money that we make here goes into that. So it’s not like the business of Mahesh is a lot of transactions. I’m a medium these days. So it comes from here, goes here.
The Bigger Mission
Mahesh: And the goal is to bring the best AI capabilities in the hands of people who are least likely to use them without help. And with that goal here, we are trying to bring a lot of people who can’t get to AI faster.
And then on the other side, we are building tools for people who can bring AI to a lot of other fields, not just tech, so that more people can come in. And if I can be a catalyst in that reaction, that’s the business of Mahesh.
Conclusion
Aakash: This was amazing, Mahesh. This was a masterclass in AI agents for PMs. Thank you so, so much.
Mahesh: Thank you for the chance. I always wanted to be here. I always looked up to you from the beginning, even when AI was not cool. I think growth was the thing. Growth was an eye-opener first time for me. Nobody talked like that about growth—it was more about building, especially in the companies I worked for.
But you brought—and I think a couple of people at that time brought—this idea to life, which is it’s an essential skill of a PM to grow the business. And I think nobody else has contributed more than you on that channel. And with AI, you are pivoting and you are iterating so fast that I can’t catch up.
So thank you for having me. I’m so humbled for giving me this experience. Thanks, Akash.
Aakash: I’m the humbled one. Thank you so much. Everyone, check out his course on Maven. Support his startup when that comes out. And we’ll see you in the next episode.
Key insights from the Mahesh Yadav podcast transcript include:
- The revolutionary “cart before the horse development” approach for AI products, where rapid prototyping leads customer discovery rather than traditional market research
- The 18-month roadmap provides a clear pathway from AI fundamentals to FAANG-level position
- Wwhile the detailed comparison of Microsoft’s innovation culture versus Amazon’s business-first mentality offers invaluable career guidance for aspiring AI product managers
