The Product Manager

How to Adapt to the New Gold Standard in Product Management (with Maryam Ashoori, VP of Product and Engineering at IBM Watsonx)

• Hannah Clark - The Product Manager

Share episode

Product management isn’t just evolving—it’s getting completely rewritten. In this episode, Hannah sits down with Dr. Maryam Ashoori, VP of Product and Engineering at IBM Watsonx, to talk about how AI is reshaping not only the tools product managers use, but the very nature of the role itself.

Maryam brings two decades of experience in AI across design, research, and engineering, and shares what she's seeing from the front lines of one of the most powerful Gen AI platforms on the planet. They explore how AI is shifting expectations, blurring team roles, and opening new paths for productivity and creativity. If you’ve ever felt like the PM playbook just got tossed out the window, you’re not alone—and this conversation is a thoughtful look at how to navigate what comes next.

Resources from this episode:

Hannah Clark:

As technology evolves, so do our behaviors. In the past, those changes would feel pretty incremental. Like instead of handing in a resume in person, now we apply online. But in a time of drastic technology changes, product people are in the middle of a total reinvention. Not only are we adapting to new behaviors as users ourselves, but we're also adapting to completely new standards in the ways that we work. And let's be honest, that shift sometimes feels like we're building the parachute after we've already jumped out of the plane. So how do you build a decent proverbial parachute when you're already disoriented and there's no formal manual to instruct you? Well, a good start is to look around at the skydivers who have managed to take the same materials you have and found a way to fly. My guest today is Dr. Maryam Ashoori, VP of Product and Engineering at IBM Watsonx. Dr.  Ashoori, who assured me I could simply call her Maryam, has worked in the AI space for over 20 years. And while there are very few people who have sat closer to the cutting edge of this technology than she has, it's only been recently that she's seen a monumental shift in the behaviors of high performing product people. So if you'd like to borrow the playbook of the product team closest to some of the most significant contributions to AI in history, this episode might interest you. Let's jump in. Oh, by the way, we hold conversations like this every week. So if this sounds interesting to you, why not subscribe? Okay, now let's jump in. Welcome back to the Product Manager podcast. I'm here today with Dr. Maryam Ashoori. She is the VP of Product and Engineering at IBM Watsonx. Maryam, I'm so excited to talk to you. How are you doing today?

Maryam Ashoori:

I'm good. How are you?

Hannah Clark:

I'm doing great. I'm so excited to get into this conversation and we'll start it off the way we always start off. Can you tell us a little bit about your background and how you got to where you are today?

Maryam Ashoori:

Absolutely. I've been working in AI for over 20 years now, across multiple disciplines. Throughout my careers, I've worked as a designer building AI driven systems with a user in mind. I've watched as an engineer bringing that data and optimization to the systems that we built. I've watched as an AI researcher pushing the boundaries of what AI could do, and ultimately I found my home in product bringing all of these perspectives together to solve the right problems. And more recently. To be precise. Over the past two years with generative AI, I've been at the forefront designing and building for what's the next study, AI, which is the core IBM Gen AI enterprise platform. From the ground up, that work brought together everything that I love from defining strategy in fast-paced market, the evolving market to navigating legal and ethics stuff, and new field like Gen AI in this case, with all the limitations associated to that. To build something useful that solves a real problem. And throughout that, I realized, and I've come to see responsible AI as the crucial element that we have to closely work on and design for. And that led me to my current role, which is leading product and engineering for AI governance.

Hannah Clark:

This is such an honor for us on the show because AI has been such a huge conversation for us, and so it's so fascinating to be able to talk to someone who's got the context of being in this field for much longer than it's been in the public conversation in the way that it is right now. But today we're gonna be focusing specifically on the role of product management and how it's evolving with the emergence of AI-powered tools and workflows. This is something that's really close to your field. So I'd love if you could kick us off by telling us a little bit about some of the shifts that you've been observing at IBM Watsonx and what those changes have been signaling about the future of the profession.

Maryam Ashoori:

Well, there are two things that I would highlight. One is on the pro profession itself, and the second one is on the products that this product managers built. Let's start with the first one. There's no doubt that there is much acceleration into the productivity aspects of Gen AI bringing to everyday life of a product manager. If I just look at my own product managers, they are white coding. They were not white coding a year ago, so in a year ago, if I wanted them to create a PRD and go work with designers to put together the mockups and then go to engineers to put together a prototype and come back, pitch the idea. Now they just build it in less than 24 hours and they show me something fully fledged, fully functional, and I'm like, what am I looking at? Is it the actual product or is the mockup that you are building? It shows how the profusion is changing using AI and at every layer, this was just acceleration to test out an idea. But you can also think about PRD generation using AI to help you brainstorm and evaluate multiple options that you have. So that's on the productivity side of the house, on the product side of the house, these product managers are assigned to build a product that solves the problem. That product, there is a very good chance that can potentially benefit from the acceleration of Gen AI. So it's essential for these product managers to have a deep expertise of how this technology can help my own product and my own work and bring those acceleration to what they are building.

Hannah Clark:

Yeah, and I think this is like a competency that is, it's a very interesting space right now'cause it's rapidly evolving. There's not really a whole lot of formal education that's able to kind of keep up with the pace of change. From what you've seen in your own team and in the field right now, what are some of the ways that product managers that are wanting to stay competitive in this field and really keep up with that productivity gap, stay competitive. How can we build up our skillset?

Maryam Ashoori:

That weakness that you mentioned represents an opportunity. All this acceleration, at some point in the future is gonna become a norm. The metaphor that I keep thinking about is calculator. Many years ago, people were doing manual calculations, and even I remember at the school calculator was banned because they wanted us to do it manually. But then we crossed that chasm and we got to a point that you just move on to solving different problems rather than doing manual calculation. It's the same metaphor and the same thinking. The product managers on the future are expected to effectively use AI in their job, but there is a period of transition that whoever takes advantage of that can potentially create an opportunity for themselves to stay ahead of the game. So this weakness is actually representing a good opportunity. And I remember recently, actually we ran a study with thousand people just focused on. Figuring out that productivity acceleration that we get from AI. And we asked them one simple question. These were thousand AI application developers focused on enterprise building in US. And then we asked them, Hey, do you use AI assisted coding? 38%? They said frequently they use that and we are like, wonderful. How much time saving are you getting out of that? And 41% they said one to two hours per day. So one to two hours of time saving per day. 4% they said more than four hours. What does it tell me? And it's independent from the road. If you know how to effectively use AI to get AI to work for you, you can potentially be unlock so many new ways of thinking, even in product management that we haven't explored there. And I think that's the true opportunity that this technology represent for product managers.

Hannah Clark:

I would agree. I think that Vibe coding in particular really combines a number of different skill sets that are kind of unique to an experienced product manager and able to kind of create something that can be a starting point that's, like you said, worlds ahead of just the PRDs that you'd be kind of starting at early on and you know, a year ago or two years ago. So now that we think about the tools that are involved in AI, speaking of vibe coding, you've noted in a past conversation that building an AI application usually requires a lot of tools, about six to 15 different tools. And so if we're working with limited AI expertise, how do you recommend approaching the challenge of understanding and orchestrating complex technology stacks, especially now that tools have become more complex and more sophisticated?

Maryam Ashoori:

I would say two things. Think about the technology part of this question and the people part of this question. If you just purely look at the technology, the field is evolving rapidly. As a product manager, probably you have less than two hours to go and explore a new thing that comes to the market to make a decision if this is something useful that you wanna spend more time and bring in to your product, or it's just basically treat it as a noise and leave it out there. So limited time evolving market, limited expertise in AI. As a product manager, you probably don't have a deep expertise in AI the same way and level that an AI researcher has spending years on this. So really. Figuring out how to evaluate the situation and the technology out there. If you just rely on the tools, I feel like that's, you are missing out on the people part that can come and help you because of the limited knowledge that you have. On the people side, when you look into how the market is evolving and how different roles are evolving in the market, probably the best help that you can get is from your partners in engineering and your partners in research, or in some companies they call them scientists, research scientist. Because they understand how the field is evolving and what is the value of that technology to what you are applying. If you look into the field as a whole, the line between these roles and responsibilities are dissolving to, in particular, between engineers and product managers. Historically, we were saying that, hey, for every one product managers, you probably wanna have five to 10 engineers. A month ago, we saw Andrew Inc on a stage and he said, Hey, my team proposed to have one product majors for every 0.5 engineer, so you have more product majors, and this is exploration. Not saying that's the right. Ratio, but what it shows is the line between roles and responsibilities are changing, so does the expertise that is needed to build the right product. My advice for product managers is to focus on building the right product because once you name that definition, you can pass it to your human partners or the AI partners to. Make them work for you, but the essence and the crucial part is to get it right and define it right.

Hannah Clark:

That is the challenge that we're all kind of trying to figure out is what's the balance between how much of the tools we want to lean on and how much of our own, how do we kind of inject our own expertise and the expertise of those that we're collaborating with. So when we think about evaluating AI technologies themselves for a product roadmap, for example, what are some of the criteria that you'd use to distinguish between tools that are worth investing in versus those that might be here and gone again in six months?

Maryam Ashoori:

Yeah, I've seen people that chase solutions versus solved problem. A very good example of that is someone that came to me and said, Hey, by the end of the year I wanna bring in two agents to my product. And I'm like, hold on, like two agents to do what? Exactly. So this to me is chasing the technology to figure out how I can bring it versus. Having a really good understanding of what is the problem that you're trying to solve, and if this new AI tool, it's a distraction or it's a benefit, true benefit to your product, I would start there because then you are gonna have a framework and lens at which you can evaluate the value of any piece of technology that comes your way. Figure out if you wanna spend more resources or more of your time on evaluating that.

Hannah Clark:

Yeah. You know, this is echoing a conversation that we had recently on a panel event. Thomas Stokes, who's a co-principal at Drill Bit Labs. So he's heavily in UX research and he was echoing some of those same sentiments in which we're kind of seeing this sort of a emergence of this. Solution first thinking of how do we use AI rather than thinking about the problem first. And I think that this is a really good moment in time that a lot of us kind of have to go back to basics and think about, you know, like the technology is very exciting. It's, there's obviously a lot of use cases for it, but we still need to be thinking about things from this problem first mindset in order to ensure that we're using it in a way that's accurate and actually solves the problem of the heart of things. So how do you, at IBM, how do you strike a balance between building AI capabilities in-house and adopting third party solutions? I know you guys are pioneers in this space, so I'm very interested to see how you guys are approaching this kind of an issue.

Maryam Ashoori:

I wouldn't say it's an issue. It's an opportunity to amplify your effort, especially when this comes to Gen AI. Just look at what's going on outside in even open source. The innovation is coming from academia, is coming from industry, it's coming from developers just doing it on their own as a passion project. It is coming from. Every resource. So if you limit yourself to, Hey, I just wanna build in-house myself, I feel like you are just restricting the access to innovation that you could possibly have. So for every single area that we focus on, usually we are looking to what is the value of leveraging our in-house capabilities. For example, for us specifically, we have access to 2,500 people in IBM research that are living and breeding the state of the art technology. So I have very frequent conversation with them to figure out where their thinking is and what can potentially benefit the product. But at the same time, I'm having the same conversation with my partners. In the market or even the community, like selected people in the community that are shaping the community to see, hey, where do you think it's going in the next three months? How can we support the community building activities? And at the same time, I'm making my time and I know that's the hard part, but I am dedicating time. In my calendar just to go and look for random news that are out there because that's where you find the passionate developers building something, putting out there, which is not in the spotlight of big companies announcing major. Features. So throughout these channels you get a visibility and exposure to broad spectrum of technology that is coming to the market. And then back to the point of what problem I'm trying to solve, then I can look around and see what is the cost and benefit of getting this piece of technology from here in. Versus community versus those developers. Maybe I can hire them or I can do something about it to bring it and then make a, an informed decision.

Hannah Clark:

Oh that's a really holistic way of looking at that kind of framework of decision making. Speaking of kind of decision making, I wanna kinda shift gears into more of a leadership perspective here. So, you know, as someone who's leading a team, what are some of the new skills and competencies that you're really looking for when you're in hiring mode for new product managers to join the team? Even compared to a year ago when the technology was newer?

Maryam Ashoori:

Again, I would look into the problem that I'm trying to solve. With that hire, I typically look into the skillset of the people on the team and identify the gap because we usually hire for that gap. Then for that gap, I would look out the bar for me is, would I one day report to this person? So if I wanna hire someone, you wanna hire someone that knows more, is filling that gap and is the expert in that very specific thing that you are hiring them for, and then get out of their way. Give them the autonomy to deliver that because they should be truly the best. So that's the mindset that I'm adopting, really understanding what are the gaps and for gaps. I'm not looking into, oh, I have a product manager, so I have a designer, no, specifically, like this person is very good at, for example, go to market and customer conversation. The other person is very good at growing the SaaS space. The third person is at what? So basically you looking into skills versus titles. Try to feel that skill and find the best person that I can possibly bring on the team to cover that.

Hannah Clark:

This is a very interesting thing because I feel like this is also a shift in thinking in terms of, you know, what are some of the tertiary experiences that this person or. More scenarios or connections or other aspects of this person who stand out versus just qualifications. There's a real departure from a different kind of hiring mentality In the past when we're really focused on very specific hard skills and certifications and qualifications, it seems now that the breadth of someone's experience and other things that they bring to the table that are unique to them as an individual matter more. Would you say that's accurate?

Maryam Ashoori:

That's accurate. And I think what's common in most of the hires that I've seen over the past two years, we, at the age of Gen AI, is basically having full characteristics. One is be curious, show that your curiosity about where the field is going, jump on it, explore, build things. And the second one is. Be technical, really understand what's going on. I've seen some of the PMs that they categorize themselves as a go to market, PM versus technical pm and I'm like, hold on. We are talking about the age of AI. You need to have a very good understanding of how it can help your product and yourself. And if you can't answer that, maybe you're in the wrong position.

Hannah Clark:

Yeah, I agree. What you'd said also about what I wanna report to this person someday really reminds me. I spoke to an Italian about Kova at Atlassian a, a year and a half ago about a similar kind of thing about things that she looks for in a candidate. And one thing that I thought really stuck out was. If I talk to this person about who they admire, have they ever talked about someone that was more junior to them? And I thought that was a really interesting thing, like looking for qualities about, you know, what is this person's ability to collaborate or be respectful to all the competencies in the chain? And I feel like now especially, is this a really good opportunity to be looking for, you know, how is this person collaborate with others, given that the technology is so collaborative in nature, I find this very fascinating. Moving on, you mentioned before that two key areas where AI impacts product management was, you know, enhancing the existing products as well as improving the personal productivity. I wanna talk a little bit now about how we enhance our existing products with AI since we've, you know, we kind of touched on this briefly. What are some of the really brilliant use cases that you've seen in both of these areas? I'm sure that you're full of examples.

Maryam Ashoori:

So there are different ways that we can bring AI to our products. One is you can be the supplier of that technology to the world. For example, what's the next AI? The platform that I build is packaging LLMs and tools and everything that you need, and offer it to other roles to build up on. So that's one way of thinking about it, bringing them as a platform, supplying them to the market to help application build applications. The second one is to build applications that are enriched with AI. So no matter what product you have, you can think of, oh, how can I bring AI to enrich some of these features that I have and take advantage of the acceleration. It can be either one of the direct use cases that gen AI unlocks, like classification information, extraction question, and answering like customer care, chatbots, basically content generation, code generation, one of them. Or it can be automation. It's like just, you know, agents. You can connect all of these use cases that I mentioned and through function, calling to every single business workflow that you have. So how can I bring that efficiency to my product to have a better product? So that's the second category. The third category is how can I come up with a new category of product that is powered up by LLMs? One example is all the code assisted technologies out there. LLM, behind the scene, and now you're using it to help the developers to do vibe coding, basically. So that's a very good example of that. New opportunity areas. And the fourth one is, I would categorize that through services. The challenge that you mentioned for. Product managers on education is not unique to this role. It's unique to the whole market. Enterprise market, consumer market. Everyone needs to educate themselves to make better AI decisions and it represents an opportunity for services. So if you know the field, you can go in and help them pick the right model for the right use case. Make a better decision. So it represents four different areas of opportunities for product managers, and depending on which area you're targeting, the advice for how to approach that is completely different.

Hannah Clark:

I'm really interested in this fourth category of the services because I have also noticed this kind of a trend towards a popularization of even product managers moving into a services or education space. I've seen just an explosion of former PMs offering Maven courses and it seems like there is a huge push towards feeding this opportunity now that especially vibe coding tools have made coding and development available to just about anyone. I can really see that trend towards, well, how do we kind of offer other services to empower people to develop the solutions that are really bespoke to them? So it's really interesting. That's a whole other shift that's kind of occurring kind of concurrently with all the developments that are happening in product, I think is really interesting. So looking ahead then, what would you say are some of the developments of trends in AI that you think will have a really significant impact on the next 12 to 18 months? Just given kind of what you're seeing starting out right now?

Maryam Ashoori:

You know, in our world, every three months is a generation, so 12, 18 months is five generation ahead. A little bit too far out where we are gonna go is too far. But if I just look into maybe three generations, six to nine months from now, I would say that. I'm gonna see a lot more on automation, and we see that today, but the technology is not there yet. It's more like exploring new use cases, figuring out collectively with the community how to effectively use this. It represents an opportunity for product managers, and when I say automation is literally automating every single thing from email generation to talk to your customers, to optimizing your roadmaps, depending on the dev effort needed to deliver something. At the mo. What timeline to generating content for your PRD. To exploring new options for your customers to help them make better decisions, like literally every single decision that you're making. I argue that it can potentially be enhanced with AI if you figure out how to make AI work for you. On the other side though, what I would add is all these beautiful acceleration that we get with AI. There is a cost associated to that, and that cost is the risks you can make or break your career depending on how you're using this. And it's essential for product managers not only to have a very good understanding of the benefits, but also risks. Because risk can impact their product. Week time things can go wrong, risk can impact their or personal identity and branding as well. So I would say that like always think about the cost benefit balance of what you're doing.

Hannah Clark:

Yeah. And well, and I'm sure that this is an issue that's so close to your heart. Given that ethical use of AI is a cornerstone of your career, how do we kind of navigate those risks? I feel that oftentimes mistakes can be made by really well intending people who are really just kind of in this play and exploration mode. So do you have any kind of framework or kind of guardrails that you like to think about when exploring with AI or anything that you would kind of impart on folks who are kind of looking to incorporate risk management into their development and their skillset?

Maryam Ashoori:

I'll give you an example. Let's say that you wanna use an agent that sends email on your behalf or summarizes your emails and it has access to your calendar to schedule something on your calendar. It's amazing the benefits that you get from them. On the other side, you have to give access to this agent to your email and your calendar schedule, knowing that potentially a bad actor can misuse the information of when you are gonna be at what time and talking to who. So every permission that you give to these agents potentially creates a vulnerability point for you, and that's not just for you. It's the same for your product when you bring those to your product. So I would start with back to the drawing board of use case. It's essential to understand what is the problem that you're trying to solve for the community, for your product. What is this thing? And then when it comes to AI, have a very good understanding of what is the benefit to that. But also what are the vulnerability points that I'm introducing to this product? Am I confident to have a mitigation plan to resolve those, or I'm just blindly following and bringing that AI to apply to my product and see what happen.

Hannah Clark:

Yeah, I'm really glad that you said that because I think that is kind of the less sexy part of AI development and exploration. This is time of such excitement when risk management, I think, has been a little bit of an afterthought, if a thought at all in some circumstances where, you know, we're sort of entering this wild west phase of development and exploration and almost like a second generation of pioneers with regards to vibe coding and using AI for completely new use cases. So I think that is kind of an important thing to call out. I did kind of want to end on sort of a, you know, next steps for those who are interested in developing their own skillset. In this current generation that we're in right now, what have you found to be some of the most useful resources, either for yourself or for folks on your team who are upskilling naturally? Like just playing out with technology is one way, but are there any other kind of resources that you'd recommend?

Maryam Ashoori:

Yeah. In addition to all of those goodies that are out there, I would also say find your way to filter out. What do you wanna be exposed to? Because there are so many things happening in the market, and I've seen product majors dealing with this fatigue of technology fatigue, maybe we call it. So where do you get your data? And I think the most effective. What I've been encouraging my teams to do too is to find your trusted voice in the community and follow them. If it's on LinkedIn, if it's somewhere else, just follow them because if they have the point of view that is consistent with what you are thinking. There's a good chance that a new piece of technology comes out there and you wanna hear the perspective from them versus going to the community and figure that out. And it's like a shortcut to hear their opinion, the opinion from an expert on a technology. And that's very helpful. But also like even given that you know, for your product managers cost and benefit, you gotta make sure that the person that you follow is a really worse following kind of person. Because otherwise you are limiting your view to the world to that. Window, and if that's not accurate, then you have a clouded view of the board. So be very selective, but find people that are sharing the same point of view and are expert in what you care about and just follow them.

Hannah Clark:

I love that advice and I think that it's very important right now as we're kind of entering a period of rapid saturation. The advice of being selective, I think applies to many things in AI and technology and I think we'll find also in life. Also to your point about trusting folks and following them, how can folks follow your work after this episode?

Maryam Ashoori:

I'm on LinkedIn and I try to write as much as possible, so I would love to stay in touch with them on LinkedIn.

Hannah Clark:

Wonderful. Well, thank you so much for joining us,  Maryam. We'll make sure to add the link to your LinkedIn in our show notes today. Thank you so much. I really appreciate your time. I know you're such a busy person.

Maryam Ashoori:

Wonderful. Thanks for having me.

Hannah Clark:

Thanks for listening in. For more great insights, how-to guides and tool reviews, subscribe to our newsletter at theproductmanager.com/subscribe. You can hear more conversations like this by subscribing to the Product Manager wherever you get your podcasts.