What we talked about
Saranyaa Parthikumar shares her journey from Java developer to MBA, then Lead Business Analyst, and finally Product Manager. She explains how the switch happened through an internal move: raising her hand, talking to managers, taking on junior PM style projects, and building confidence by representing the customer and guiding delivery.
Show notes
Saranyaa Parthikumar hadn’t written code in 12 years when she started using AI-assisted coding tools:and found that watching the tool explain which JavaScript file it was editing and why made it feel like a peer rather than a black box. That shift in how a non-coder can engage with engineering work is central to how she thinks about what AI is actually changing in product teams.
What we covered
- In the last eight months Saranyaa has been building working prototypes before writing a single requirements document: using tools like Lovable and Cursor to create an interactive prototype first, then using that artifact to align designers and engineers, which she says has meaningfully increased team velocity and produced crisper documentation.
- She describes zero-to-one product work as detective work:you receive an ambiguous one-sentence problem, collect clues from data and users, run role-plays (which are prototypes), and assemble a 360-degree strategy narrative before marshaling the team for execution.
- AI has changed performance expectations in a specific way: where managers previously asked “how did you execute?”, they now ask only “what did you achieve?”:the assumption being that AI removes all execution constraints and infinite resources are available, so the only conversation is about outcomes.
- The most over-hyped AI claim, in her view, is that it will replace humans wholesale. LLMs are exceptional at processing existing knowledge:she compares them to going from one scholar to 10 PhDs to 100 PhDs:but they cannot do “creative decisioning,” like a mother coming up with 10 different games to pacify a crying child.
- On trust as the primary metric for any AI feature: she points to the way Alexa can resurface a question you asked two months ago and combine it with new context in a way that makes users wonder if it was secretly listening:that feeling of undisclosed behavior is what destroys product trust.
- Her advice to aspiring PMs: technical degrees are not required. The foundational skill is the ability to put yourself in someone else’s shoes and identify their pain. Build small side projects using generative AI tools now:a resume writer, a photo enhancer, anything:and use the experience of defining requirements, getting feedback, and iterating to demonstrate PM competency.
About Saranyaa Parthikumar
Saranyaa Parthikumar is a product leader working at the intersection of product management and AI, with a background spanning Java development, an MBA, and lead business analyst roles. She has specialized in B2C products and data-heavy platforms, including growth PM work and global platform development.
- LinkedIn: http://www.linkedin.com/in/saranyaap
- Website: https://amazon.com
Episode 110 of the PreVetted Podcast.
Full transcript
Federico Ramallo (00:00) Welcome back to the pre-vered podcast where we spotlight extraordinary people and remarkable talent reshaping our world. Today we are joined by Saranya Parthikumar. Is that correct? Parthikumar? Saranya Parthikumar, a product leader working at the intersection of product management and AI. We’ll talk about her journey into product.
Saranyaa Parthikumar (00:15) That’s correct. Yes.
Federico Ramallo (00:27) what she’s learning as AI changes how products are built and the practical habits she uses to shift value without losing the human side of the work. Saranya, welcome to the show.
Saranyaa Parthikumar (00:39) Thank you Federico, thanks for having me.
Federico Ramallo (00:41) So let’s start talking about how did you move from lead business analyst to product manager.
Saranyaa Parthikumar (00:48) Yeah, my journey had multiple different hops. After my undergraduate, I worked as a Java developer, and then I went on to do my MBA and switched into the business side of things. And this lead business analyst was a culmination where I could use both my business skills as well as the technical skills. And when I was a business analyst, that was the time when product management was like a new
career option that was coming up and it was very clear to me that product managers make the call on behalf of the customers. So they needed somebody with the business sense, but also who understands technology and can talk to the technical teams in their language. So that is what prompted me. And initially it was an internal move. I spoke to the same company. I spoke to the managers. told them, hey, I’m interested.
interested
in exploring. So if there are opportunities where you will be looking for a junior product manager, why don’t you consider me in? And I got a couple of projects where I was able to represent the product manager role. And once that was done, I was very confident in the skills that I acquired. I was able to showcase my experience, and it helped me jump into a role with a product manager title.
Federico Ramallo (02:04) Amazing, And how did your background influence the kind of problems you want to solve as a product manager?
Saranyaa Parthikumar (02:12) Definitely we shine where we are more comfortable. So in my experience, I’ve predominantly worked on products that are either B2C or where there is a large data component involved. So data platforms or like a growth product manager role where you are consistently looking at data and then making iterative experiments. B2C products where you need to pour into the research report.
look at segmentation information, look at A-B test results. So wherever there is a combination of technology and numbers, I think that is where I shine. So predominantly those are the different types of products and challenges I work on, either enriching the experience, redesigning something new, looking at the data, creating global platforms, know, big data and millions of data sets. That’s how I’ve charted my path.
Federico Ramallo (03:07) Very interesting. And what was the biggest change in how you worked?
Saranyaa Parthikumar (03:11) before and after a product manager, think.
Mostly, once you become a product manager, you kind of lose control of your calendar and schedule. So that was one big change. I don’t know, like you’ve been an engineer, you’ve been a PM and a CEO, so maybe you can appreciate it better. So when you are an engineer, you are in meetings either to understand the requirement or in meetings where you divide up the task. But as a product manager, you need to be everywhere. Any context is golden.
Federico Ramallo (03:20) you
Saranyaa Parthikumar (03:40) So you need to attend multiple meetings and you are also influencing a lot of different cross-functional stakeholders. So you need to influence design, need to influence research, you need to influence engineering, you need to influence different verticals inside the organization, you need to influence marketing. So there are…
constant context change. So in one meeting, you’re talking with the designer. In the next meeting, you’re talking with your go-to-market partner. And then the second one is you are expected to be in all of the meetings to help the team march as one unit. So you lose control of your schedule. So I think that was the biggest change where you have authority, but at the same time, you have tons of responsibilities as well.
Federico Ramallo (04:29) Interesting. Yes, because now you have to interact with so many different people and you have to influence everybody to get consensus on what you’re going to build next, right?
Saranyaa Parthikumar (04:41) both consensus and also transparent communication. You become that middleman or that peg which is percolating information in all directions so that the entire unit is marching in one direction.
Federico Ramallo (04:54) Right, amazing. And what problem do you most enjoy solving as a product manager?
Saranyaa Parthikumar (05:00) It varies.
In the life cycle of a product, if you look at it, initially the zero to one, they say, go find a problem for the customer that the customer doesn’t even know exists. And then is this problem big enough so that you can have a considerable market share? And then as the product launches and is stable, then you are like, this is the cash cow. Make sure the engagement doesn’t fall. And then in the late stage of a product, you are thinking, how do I pivot this product?
I’ve kind of seen all of the products, I’ve predominantly worked in 0 to 1 and then the stable phase. So 0 to 1.
It is very interesting because there is a lot of ambiguity initially and you are kind of this detective who takes that one sentence question and then you are supposed to look for clues and trying to pitch in like what happens and come up with a story or a narrative. think zero to one products are similar. So you get like an ambiguous problem and then you put on your detective hat. You try to collect information. You try to collect evidence.
you try to do a role play which is nothing but prototyping or doing an MLP. And then you try to come up with the 360 story as to what is the strategy, what is the market share, what is the business angle there. And then you marshal the troops for execution. So I think I love it the most.
Federico Ramallo (06:26) Right, right, that’s amazing.
So how do you explain good product management?
Saranyaa Parthikumar (06:32) Yeah, I mean,
So in my mind, product management as a profession has been around for maybe 25 years. And I tried to find the history. Whatever I gathered was initially product managers used to be in the consumer marketing companies. So think Procter and Gamble or like Unilever, Coca-Cola, Pepsi, those kinds of companies where product meant a physical product. Like you could be a product manager for Lays or you could be a product manager for Pepsi.
and then you are thinking about that product in entirety like how do I market this? How do I stock it? How do I anticipate what is the number of units that need to be produced? How do I make sure people still like the taste? How do I make sure what are the different flavors to produce? And from marketing…
kind of it seems to have hopped on to technology. So very similar as a good product manager, you need to put yourself in the user’s shoes. You are like, okay, if I am using something, this product, if I am a user of this product, what are the challenges it is solving for me? So for example, now with Alexa, we say, hey, Alexa should be the world’s best personal assistant. So if I am
the consumer, what will my personal assistant do? What are the pain points that I’m facing when I don’t have a personal assistant? What are the pain points I’m facing by using like scheduling tools or something else that doesn’t understand me? How much time am I spending there? What will be my biggest unlock? So think from the customers shoes or the user shoes and then you translate it into
what can the technology solve for you? So that is how I define good product management. Start with the customer first and then translate it into technology so that at the end you get something that is achievable and something that you can define to your cross-functional teams.
Federico Ramallo (08:28) Right, right, interesting. And when did AI become part of your product work in a serious way?
Saranyaa Parthikumar (08:35) Maybe in the last year or eight months or so. So when like ChargePT launched there was an overall buzz but nobody knew how it is going to impact. But in the last eight months or so I think there have been tons of new tools. So for example
As a product manager, am creating a lot more prototypes, which are like working prototypes today. So if I’m pitching a particular idea, it is very easy for me to create a web app. It’s very easy for me to use tools that will create a UI for me. You know, you could use lovable. You could use any like ID, something like cursor, replet, all of it like supports now. So you go there, you define the
you define the interactions, you can define the data and schema as well if you are very technical and then it creates a working prototype which makes it very easy for you to visualize and also play with it. So I think in the last eight months or so
For every product idea that I pitch before I write the strategy document, before I write the BRD, before I write the use flow document, I am creating these prototypes using the GenAI tools. And then I’m using that to dictate how my thinking as well as how my requirements capturing process happens. And it’s been very effective in the sense when you are thinking
When you are visualizing a product, typically all of these requirement documents, are very lengthy and elaborate because you are thinking about like each aspect of the product. And sometimes for people who are reading it, it is difficult to connect line number one and line number 10. But when I create this prototype, it is very easy for an engineer who’s working or for a designer who needs to come up with a UI to understand what was my thinking.
in terms of the interaction or the capabilities of this product. And then they question, they add their insights, they add their nuances, and it is definitely increasing the velocity with which the team operates. So.
We are getting feedback early. We are getting a full visual solution. We are able to increase the velocity of the team. And we are able to create crisp documentation because we know what we are focusing on. So I think it’s been going great till now.
Federico Ramallo (11:02) Right, right. Yeah, I’ve seen so many interesting tools that has been able to create very unique products, things that we were unable to do before. Non-technical people can now build working prototypes and technical people can get more into the product manager role up to a certain point. So those roles are kind of merging together.
Saranyaa Parthikumar (11:21) Mm-hmm. Mm-hmm. Definitely.
Yes, yes, I think this kind of breaks the barrier between the core technical rules, right? Like earlier, if you are an engineer, you need like a particular skill set. If you are a product manager, you need a particular skill set. If you are like a product marketing manager, you need a particular skill set. And then in terms of product development, companies, most of the companies created these squads where
there’ll be one product person, a bunch of engineers, one designer, one researcher, maybe one analyst. And now I think these squads will become very fluid. They wouldn’t be needing one analyst per se. If you have a team member, even if it is an engineer who has an analytical mindset and you have a generic tool which can look at all your data set and pull you the information or let you create dashboards, you just like let that person assume the role.
for that sprint right so
An engineer can assume a product manager’s hat. A product manager can assume a BI analyst’s hat. So I think these squads will start morphing where there is some fluidity between what is the ownership and what is the functional roles each person will be doing. But then these smaller squads will remain cohesive and they’ll bounce from one problem to another. That is how I think this is going to evolve.
Federico Ramallo (12:49) Interesting, interesting, yeah. What is one AI idea you think is overhyped right now?
Saranyaa Parthikumar (12:54) There are many, but I think the most fearful one is that AI is coming for all our jobs and AI is going to replace humans. mean, initially there was a lot of buzz about.
But what I’ve realized is a lot more factors play into when AI can replace a human and in what capacity can AI replace a human. Like for example, these large language models and then the GenAI tools, they are very good in crunching information that is already available. for example, think before internet. If you were in a particular city, if you need to get an information,
you either go and talk to scholars. Next step you go and pour into the books that are available in the library and then after the internet era you started searching the internet and then after the internet boom internet became more than whatever the scholar can give you. Like I’m going and asking this researcher or like PhD some information and then the internet can give me information of 10 PhDs.
That is how the internet evolution worked. now with LLMs, think LLM can give you the information of 100 PhDs, but then it can only give information that is already present for like out of the box thinking for some of the things that require decision making with creativity. I’m not talking about generating a video or generating an ad campaign, but think where a creative decisioning is needed. For example, a child is crying.
and then a mother can come up with 10 different games to pacify the child. So similarly, will an LLM come up with 10 different things? Probably not. So I think there are limitations. The full gen AI hasn’t been achieved yet, the AI is an independent entity. It is thinking for itself, and then it can help you out.
the costs are exorbitant and I don’t know if even if we achieve it outside of the research capability, if it will be available for common use. So going with this, think the most over height is AI will replace humans. I think it can replace a lot of laborious tasks, a lot of manual intensive efforts, automation, it is going to be great for automation, but I don’t think it can replace the human yet. We will still need humans in the loop.
and we will still rely on humans for the out of the box thinking and directing the LLMs, like prompting the LLMs, giving it guardrails, directing the LLMs. I think humans will still be needed. So this seems like a pivot where a new technology is disrupting how the human society evolves, but it doesn’t look like a risk where it is going to eliminate the humans.
Federico Ramallo (15:42) Right, right. Yeah, I think that this idea of AI is taking people’s jobs is an exaggeration. I do think that with every revolution, there’s going to be, you know, some jobs are going to get lost or become obsolete, but ⁓ those people would need to adapt and learn new skills, right? But that’s a very specific subset.
Saranyaa Parthikumar (15:57) Mm-hmm.
Yeah.
Federico Ramallo (16:06) jobs, right? The other jobs are going to be requiring the use of the knowledge of AI, right? So I think that a product owner, a product manager, knowing AI is going to replace a product owner that doesn’t.
Saranyaa Parthikumar (16:20) That is true. That is true. And I think that will be true for all functions, right? An engineer who knows how to use these AI tools and increase the velocity of product making will be preferred over someone who doesn’t. A product manager who knows how to work with AI will be preferred over someone who doesn’t. But I think upskilling will also not be too much of a complicated task. Yes, it will require some kind of upskilling, but I think it is possible.
Federico Ramallo (16:49) Right. Yeah, and then the other thing is that AI doesn’t know how to abstract or understand problems. can, as you said, this can process a lot of data. So maybe it’s better at looking patterns that we don’t see, but then it is our job to, it is still on the human side to understand and give it direction, right?
Saranyaa Parthikumar (16:58) Yes.
Mm.
Definitely, definitely. Yes, I think it’s great at automation, but then it cannot determine what should be done next, right? Like for example, you give it a complicated problem, it still struggles. So we’ll see.
Federico Ramallo (17:25) So how has AI changed what you expect from engineers, designers, and also your peers?
Saranyaa Parthikumar (17:32) What all I think of?
The greatest change in expectation has been the speed of delivery. So earlier, if there is an engineer and you will be like costing a particular change with them and then you will be like, what is the level of effort? When can we get this? And is it going to impact? Same with design, we’ll be like, okay, this is the interaction. Now, like, do you need two weeks to get back to me with an initial draft of the design specification? But now people are like, okay, I’ve told you
all this, record this, here is the AI summary of the meeting, now go and get it to me in the next one hour. So I think that has been the greatest change in expectation where everybody is expecting instant output.
And then the second one I feel is slowly the evaluations are changing. Like earlier, if you are a great product manager, you will be evaluated on your outcome.
as well as your execution capabilities. know, like how many did you launch? What was the satisfaction between the teams and customers on what you launched? And then like, is the adoption and engagement of what you launched? But now I think it will be mostly on the outcome. So.
What did you do? How much did you launch? Because people assume that there are no blockers for the execution and that and you have like infinite resources at your disposal because AI can do infinite things. So that is the expectation. they are like, did you achieve more with limited resources? That conversation has changed into what did you achieve? No one is talking about the limitation of resources. So I think the speed of delivery, definitely everybody is expecting it to become much, much faster.
Thank
Federico Ramallo (19:18) Right, right. And we talk a little bit about the merge of core roles, right, ⁓ which is also happening, right.
Saranyaa Parthikumar (19:23) Mm-hmm.
Yeah, I mean,
initially, I’ll tell you my personal experience. Like, I haven’t done coding for maybe the last 12 years. And then initially, when I was doing this wipe coding, I was a little jittery, like.
would I have to go back to the code, look at it, and then make fixes? But then it is so natural in your natural language. And sometimes when it says, I’m looking at this JavaScript file, and then I insert a comment in this paragraph, and then before this comment, I will add this because you asked me to change the feature.
It isn’t scary anymore, right? Because you are not doing it yourself, but then you understand what the JNI tool is thinking and what it is trying to accomplish. So it kind of feels like another person or another peer who is doing similar things, but at a faster speed. So yeah. Sorry.
Federico Ramallo (20:19) Right, right.
Very interesting. I think we talk a little bit about this, but what new skills do you think PMs need most to learn because of AI?
Saranyaa Parthikumar (20:29) The first one I would say is AI gives you infinite resources at your disposal. So PMs need to be very, very strong in prioritizing what are the biggest pain points they need to go after. And then the second one I think is
It is very dynamic in how it behaves for every scenario. So you need to have rigorous testing to understand
if ethically and then safety aspects of it, the factual aspects of it, like if it is not hallucinating, it is giving you the correct answer. So you need to do a very, very rigorous testing to understand if the product meets your requirements. So I think these are the biggest changes.
Federico Ramallo (21:17) Right.
How do you decide what not to build when AI makes everything feels possible?
Saranyaa Parthikumar (21:24) Again, a difficult question, right? So very difficult, but I think you need to think in terms of customer journeys. So like I mentioned, put yourself in the shoes of the customer and then see if it makes sense. Like for example, we can create automation tools that will probably generate
Federico Ramallo (21:26) You
Saranyaa Parthikumar (21:45) image every day like a good morning message every day that you can blast in your social media apps versus a productivity where it is summarizing what are your important emails what are your important messages so that in the morning you are not like frantically checking your phone for 30 minutes figuring out what did I miss in the world
It depends on as a product manager, it depends on who your customer segment is and then what is the benefit you are willing to give your customer. So if the benefit you are willing to your customer, like I said, Alexa is a personal assistant. So as a personal assistant, I would rather focus on, hey, I’ll summarize your day so that you’re not frantically checking all your applications for 30 minutes, figuring out, how should I plan my day today versus I’m Alexa.
I’ll help you create like 10 different pictures and 10 different DPs that you can put on your social app every day. It’s possible. Both are possible. Probably depending on like who’s using the personal assistant, both might be important for a particular person. But as a product manager, I’m thinking, the productivity use case might be 60%, 70 % of my customer segment. And that will have a higher impact and improve the quality of their life.
or then I am giving like 10 different profile pictures and 10 different good morning messages, right? So I think the prioritization has to come from understanding the customer’s life and then determining which one has the highest impact rather than what is technically possible using AI.
Federico Ramallo (23:21) So how do you make sure AI features are trustworthy for users?
Saranyaa Parthikumar (23:26) Yep, so probably I would add that understanding how to evaluate AI is another skill that all product managers should learn going forward. Definitely, you know.
Looking at the AI responses and then comparing to what a good response will look like. Having a rigorous human in the loop testing before you launch your product. And the ethical considerations.
Should I allow a kid to ask a question? Should this feature be available to, you know, like so and so if they ask a particular question, like for example, someone asks, how do I build a bomb? And then like, do you answer it? you say,
No, I cannot. Or do you say, why are you asking about this? Can I help you? So I think there are like ethical and then sensitivity aspects and the correctness aspects. So first one will be you need to vet all the AI responses, have a human in the loop, have a rigorous testing. And then the second one would be like, make sure all your guardrails are up and then you are good to go. And I’m talking about only the applied AI aspect of it.
if you are a core AI manager and you’re working on a model, there are like different aspects that you would look at in terms of how is your model better than the other, what is the speed, you know, and what is the amount of challenges that you can accomplish, so on and so forth. But for somebody who’s using AI and applying it to solve problems of real users, I think whatever I mentioned in
of rigorous testing and then the guardrails and sensitivity, those are the two important factors.
Federico Ramallo (25:06) Right, Yeah, at the end of the day, it’s that way we keep circling around the importance of the humans, right? Because it’s the human criteria, the one that is going to set the tone of whatever we’re going to build, right?
Saranyaa Parthikumar (25:20) Absolutely.
Federico Ramallo (25:21) What metrics matter most to you when launching an AI feature?
Saranyaa Parthikumar (25:26) I think the primary one is trust. know, like if I compare it to Alexa, a lot of the time people are like,
Hey, like is Alexa, you know, like secretly recording my conversations. And sometimes, you know, like you don’t realize what is the context window for like, you know, like how, what is the memory in which all these contexts are stored by the AI. Like for example, I would have asked AI about Super Bowl like two months back. Oh, is Super Bowl happening in California this year? And then I might have forgotten all about it.
And when I ask Alexa about…
What is happening with Superbowl? Alexa might combine it with whatever I asked two months back and say, by the way, you asked me if this happening in California. We spoke about it. And then this is this. I wonder, was Alexa listening to all the conversations I have with my family? How did Alexa know about this? Our short AI memory is long. So I think the main factor for any AI feature is trust.
the humans feel you know there isn’t enough disclosure this AI is doing things that it did not tell me it is going to do or they feel threatened they’ll stop using the product or you know like for the whole industry itself it will go into a wrong turn so getting customer trust will be the most important metric I will look at for any AI feature
Federico Ramallo (26:59) Right, right, very interesting. What advice would you give to someone trying to start a product management career today, especially how AI is changing this role?
Saranyaa Parthikumar (27:13) I would say don’t hesitate. There has been like lot of rumors out there in the open, but it is the reality isn’t as bad and as a product manager.
Typically, you are expected to have very good soft skills as well. You are expected to be a great communicator. You are expected to be a great collaborator. You are expected to be a great team player. None of that has changed. So for somebody who’s looking to branch into product management, I would still say choose a sweet spot where you are more comfortable in. If you are thrown an ambiguous question, can you come up with five different ideas for solutions?
and then see if you are comfortable in working in a collaborative environment and then the technical skills will follow. So anyone who’s trying to start as a product manager, the advice I would give is…
use these generate tools, generate some of the site projects. It could be anything. It could be a resume writer. It could be a photo enhancer. It could be that profile picture generator, anything. Go build it. And that experience will teach you a lot of things like how did you interact with the tool? How do you give requirements? How do you determine what are the customer pain points? How do you get customer feedback? How did you iterate on your solutions? So now there is no cost for
building all these small small site projects. Go build it and then let people play with it, test it, give you feedback, gather it and these are going to be great experiences and then that will help you hone your core skills that are needed for the product manager and then you can pitch in yourself for a role that matches your criteria. So that’s what I would say. Go, don’t hesitate, try out new things.
you have tried out new things go and pitch yourself for the right role.
Federico Ramallo (29:01) Amazing, Sandreya, we’re running out of time, but I truly appreciated you being here today. We learned a lot about how to become an amazing product manager. Any final remarks before we wrap it up?
Saranyaa Parthikumar (29:15) One more maybe question that is in everybody’s mind. So typically people are like, hey, it’s a product manager technical rule. Do I need a technical degree or an engineering degree to become a product manager? I would say that is a big no.
People with all backgrounds can become a product manager even for a technical product. So don’t let that hold you back. If you are able to put yourself in the shoes of another person and think about their pain points, that is the biggest skill you need to have. And then everything else you can learn on the job. So that is my parting thought.
Federico Ramallo (29:50) Great. Amazing, Sanreya.
Thank you very much for joining us today.
Saranyaa Parthikumar (29:56) Thank you, thanks for this opportunity, Federico.