Welcome to The Boston Product Podcast, a brand new podcast powered by VentureFizz.
A couple of years ago, we started a Product Management Community to provide more programming and networking opportunities for Product Managers in the Boston area. Since then, we’ve grown to 300+ members on Slack and hosted a continuous lineup of inperson & online events.
Why launch a podcast? Two big reasons:
- Spotlighting Local Talent: Boston is home to world-class product leaders. This show gives us a platform to spotlight their work and dive deep into their real-world playbooks and hard-earned lessons.
- It’s in Our DNA: Podcasting comes second nature to us after running The VentureFizz Podcast for the past 8 years.
However, I wanted this show to be hosted by active product practitioners—leaders with the credibility to unpack real tactics around building 0-to-1 products, navigating the AI era, scaling cross-functional teams, and mastering the modern PM craft.
As the community grew, two exceptional product leaders stepped up: Kaitlyn Kirkaldy (VP, Product at Candex with prior experience at Zipcar) and Adam Piel (Sr. PM at AcuityMD with prior expereince at Eventbrite, Spotify, and Pivotal). After seeing them host happy hours and engage with the community, asking them to co-host this show was a complete no-brainer. Needless to say, I was thrilled when they said yes—they are natural hosts!
In this launch episode, Kaitlyn and Adam sit down with Emily Levada – VP, Consumer Product at Babylist.
Emily shares her journey through the Boston product ecosystem and unpacks how Babylist stays ahead of the curve in consumer AI, 0-to-1 product development, and so much more.
Key Topics
- Career transition from theater to product management
- The parallels between theater production and agile software development
- The role of AI and ML in modern product management
- Building a culture of experimentation and rapid iteration
- The importance of judgment and customer understanding in AI-driven product development
- The evolution of Babylist and its focus on serving parents and families
- The strategic use of AI tools like Claude and ChatGPT in product workflows
- Challenges and opportunities in integrating AI into organizational processes
Subscribe to The Boston Product Podcast on:
Connect with the Community:
- 🌐 Join 300+ PMs in our Slack community: venturefizz.com/communities
- 💼 Follow VentureFizz on LinkedIn for keeping posted on upcoming events, networking, content, and more.
Transcript
Here is the cleaned-up transcript. All stutters, false starts, filler words (like “um,” “like,” “you know”), repeated words, and broken sentence structures have been edited for clarity and readability, while keeping 100% of the original content, tone, and context intact.
Emily Levada (00:10) In February, our CEO said, “Stop what you’re doing. Don’t work the way that you’ve been working. Everybody gets repo access. Everybody needs to be building.”
Adam Piel (01:49) We want to welcome Emily Levada for joining us today. Hi there.
Emily Levada (01:54) Hi, it’s lovely to be here.
Adam Piel (01:56) Thank you so much for being here. My name is Adam Piel. Kaitlyn, do you want to introduce yourself?
Kaitlyn Kirkaldy (02:01) Hi, sure! I’m Kaitlyn Kirkaldy. Super excited to dive into things with you today, Emily. We’ve got a ton of stuff we want to hear from you on.
Adam Piel (02:08) Great! So maybe we can just start with a super general, cocktail-party small-talk question: give us a sense of your career so far—who you are, where you’re from, where you got started, and an overview of how you got to where you are today.
Emily Levada (02:23) I’m from Connecticut originally. I have college degrees in psychology and theater, which I like to say are both studies of how individuals behave—one artistic and one scientific. I had an early career in professional theater production, which we can talk about if you want.
I wanted to work in technology, so I got an MBA and ended up at a not very well-known company called CSN Stores, which—as those of you in Boston probably know—became Wayfair. Actually, the Wayfair brand launched just before I joined. I was there for about nine years all over the business. I worked for four different members of the C-suite across supply chain operations, merchandising, the core product experience, and then my last 18 months running product and design for the global large-parcel transportation network.
I left there to join a company called Embark as Chief Product Officer. Embark does genetic testing for dogs—think 23andMe for your dog. I was there as CPO for about two years and then as interim co-CEO for about nine months. After that, I took some time off. I did some work with ezCater—another great Boston company—and then joined Babylist a little over two years ago.
Adam Piel (03:45) That’s quite a story. I love how no PM stories are quite the same, and they often have little detours into other industries. I wrote this down to ask you later, but because you mentioned it, I’ll jump in now: can you say a little more about how that other career experience informs your lens on product management?
Emily Levada (04:04) Yeah, totally. There’s actually a tremendous amount about theater production that is very similar to product management. I didn’t really know it at the time. I knew that I wanted to go into technology; I’d kind of always been a technophile. I went to boarding school for high school, and I was what was called a Resident Technical Assistant—which was basically when your computer wasn’t working, they’d send kids in to try and fix it before IT had to come help. I took HTML classes in college and picked up SQL while doing a ticketing database migration at the theater from tickets.com to another software. There’s lighting, sound, and video technology—there’s a lot of technology inside of a theater.
I had an amazing career coach during my MBA who had a wonderful philosophy. He worked with you to help identify tensions you felt when thinking about what role was best for you going forward. As an example, two that jumped out to me were:
- I really like things that are both artistic and technical. Having a job where I could do one but not the other often felt in tension, because jobs usually had one or the other. When you feel that, it means you need to find a job that offers both.
- I really wanted to build. I felt like I would not be happy as a consultant; I wanted to build things. But I also love constantly solving new problems. Solving the same problem over and over again isn’t super fun for me.
I was trying to figure out what career would satisfy those tensions. Someone recommended a book called Artful Making. It basically argues that agile software development is more like theatrical management than anything else. You have cross-functional teams of creatives and technicians. You have an ensemble that works both backstage and onstage to create the show. Rehearsal is just an iterative prototyping process where you figure out what you’re building, and then you go from a development environment in rehearsal into a high-precision production environment. The only real difference is that your audience shows up and claps for you.
There was a First Round Review article years ago titled something like “Why Chefs and Soldiers Make Great Product Managers,” and everything in that article applies to theater professionals too. They value preparation, think on their feet, make quick decisions, and develop a “show must go on” attitude.
Careers are not straight lines. It’s easy to look at the resume of a successful person and only see success, but I got rejected from every MBA program I applied to except the one I attended. I graduated from my MBA program without a job and took a role in inventory planning, not product management. So it was not a straight line, and it was not all success. But like many product managers of a certain era, product management wasn’t a formal track straight out of college—you had to find your way into it.
Kaitlyn Kirkaldy (07:23) That’s great! The non-linear journey into product—with its detours, twists, turns, and sabbatical side quests—is so common. Speaking as someone who has had several in their career, one takeaway I’d love to implement is that maybe we should applaud product releases and go all-in on the theater model! Every product manager needs a win, and a little bit more applause would make a sometimes thankless job a bit more fun.
What I really want to dig into next is the latest chapter of your story at Babylist. Talk to us about how you found that opportunity, what was so interesting about it, and what drew you to the role. A lot of folks are very familiar with Babylist, but many probably aren’t, so it would be great to hear about the magic behind it.
Emily Levada (08:22) I didn’t know about Babylist as a consumer initially. I knew about them because their Chief Growth Officer at the time was an MBA classmate of mine, so I had been following her journey on LinkedIn. Like most people whose kids are a bit older—my youngest is eight—I sort of missed Babylist before it became the powerhouse it is today.
For those unfamiliar, more than 20% of families having a baby in the US use our services, and nearly 40% of first-time parents use us for their baby registry. More than 10 million people a year shop with us. In 2025, our GMV was $1.4 billion with over $750 million in revenue, so it is not a small business.
What made it interesting for me was stepping back after a long job search. It was hard at first to understand the common thread among the roles I was getting excited about:
- Wayfair served homeowners.
- Embark served pet owners.
- I looked at a fintech company doing student loan financing.
- I talked to a company focused on internet safety for teens.
When I stepped back and asked what all those had in common, the answer was the customer: they are all companies serving parents and families. At some point, I drew a line in the sand to focus strictly on companies serving parents and families. That was partly a requirement of the job market. I had been successful for a long time as a generalist, but the job market has changed. The more specific you can get about what you’re looking for, the stronger you are as a candidate.
Babylist had a Director-level product role open at the time. While I wasn’t necessarily looking for a Director-level role, I knew they were building out leadership on their team. I reached out to my contact and asked to connect with their Chief Product Officer to discuss what they needed as an organization, and we went from there.
Kaitlyn Kirkaldy (11:01) That’s great! Looking back at those through-lines—even if they weren’t obvious in the moment—is something everyone should do periodically. I love hearing how people weave that tapestry together regardless of industry.
Emily Levada (11:27) Absolutely! You might ask what a homeowner and a pet owner have in common, and on the surface, it might not seem like much. But the reality is the parent—a mom or dad making those decisions for the home or their pets.
Kaitlyn Kirkaldy (11:38) Right, it’s an asset you’re taking care of and love more than anyone could imagine. I totally get it. With Babylist’s massive growth, how have you experienced its evolution from an e-commerce registry solution into the multi-product world it is in today?
Emily Levada (12:30) There is no more powerful economic force than a new baby! Parents spend around $28,000 in the first year alone—it’s a hundreds-of-billions-of-dollars market. It’s a time when parents are making brand-new decisions, building brand affinity, and asking endless questions.
The registry is a tool, but what we really care about is meeting parents in this “high-importance, low-knowledge” window of their lives. We give them a tool that helps them feel confident and ready to bring home their baby.
There are so many of those high-importance, low-knowledge windows—whether that’s figuring out how to use health insurance benefits, planning for financial futures, and beyond. It makes complete sense for us to continue expanding how we serve customers across different problems and life stages.
Kaitlyn Kirkaldy (13:51) I love that phrase: “high importance, low knowledge.” That is a fascinating moment to meet a consumer. I imagine there is a lot you’re thinking about with AI in that context.
Adam Piel (14:05) You knew it was coming—every product conversation has to touch on AI! What has changed in the last year or so, and what hasn’t changed at all?
Emily Levada (14:43) How we work has changed, is changing, and will continue to change. As soon as you think you have something figured out, it shifts again.
However, the core customer hasn’t changed, and having a baby hasn’t changed. There is only so much AI can do to help you take care of a baby. When you have a baby, you need physical goods, and part of our job is helping you afford and acquire those things. We are also an inherently social product. A registry isn’t just a list; it’s an invitation for your community to support you.
Where we are utilizing AI is in expanding product recommendations by leveraging ML and generative AI to better understand our customers, catalog products across the web, and provide human-advisor-level consulting at scale. So while our tools to do work faster and more efficiently have evolved, the core customer need remains constant.
Adam Piel (16:28) How do you think about org structure around AI? Are PMs split into dedicated AI teams, or is every PM effectively an “AI Product Manager” now?
Emily Levada (16:56) We have two specialized teams: one building our product-recommendation chatbot and another building our customer-service chatbot. Those are the teams most likely to have “AI” in their title.
However, every PM is empowered to utilize LLM-based features. For example, a team building a “getting started” quiz placed free-text fields in the UI and leveraged LLMs on the backend—they aren’t a dedicated AI team, but they’re using the tech.
Beyond feature development, our whole team is moving quickly toward becoming “AI builders.” Everyone has code base access and direct Snowflake access. PMs and designers are writing and pushing code, while engineers are encouraged to own strategic product thinking end-to-end. We’re breaking down traditional functional silos.
Adam Piel (18:27) How is the day-to-day bureaucracy changing? Product orgs have historically generated mountains of documentation. How do you handle information flow now that people can generate or summarize ten-page docs with AI?
Emily Levada (19:18) Babylist generally isn’t a bureaucratic culture, which gives us a great starting point. Everyone uses AI, but there is an explicit expectation: you are responsible and accountable for everything that leaves your desk.
We use AI heavily for personal productivity. We have a shared internal plugin library, and we use AI tools for prototyping, writing PRDs, and generating weekly updates.
However, there’s an important distinction: if you only use automation to push information around, you aren’t creating true leverage. That information still has to be read, processed, and passed up and down decision chains. I’m much more interested in using AI to empower front-line teams to answer questions and make decisions themselves.
For instance, I’ve been working on an internal plugin that records product context—capturing why past decisions were made, our unstated surface principles, what we’ve previously tried, and potential opportunity areas. It acts as a first-pass sounding board for anyone asking, “Is this a good idea?” or “What feedback would Emily give me on this?”
Kaitlyn Kirkaldy (21:36) I love that! We call that codifying “product configs.” How did you build that context agent, and how did you decide what information to arm it with?
Emily Levada (22:21)
We used a two-step process:
- Data Mining: We pointed Claude at our repository to read and catalog every test report, PRD, and Slack channel on a given surface area to consolidate what we currently know.
- AI-Led Interviews: We had Claude generate interview prompts and actually interview product team members about why they made specific trade-offs or decisions.
That second step was key to capturing the implicit knowledge inside people’s heads that was never formally written down. Doing that across our core surface areas consolidated empirical data and captured underlying team rationale.
Kaitlyn Kirkaldy (24:00) That is brilliant! Capturing what lived in someone’s head back in 2017 is notoriously difficult. Why did you choose Claude specifically?
Emily Levada (24:40) Mostly because Claude Code was where I spent most of my time, so it was my comfortable default. At work, I primarily use Claude.
Personally, I use Claude for work and OpenAI at home. Logging my work account into the Claude phone app and my personal account into ChatGPT helps keep my data separated while allowing me to stay proficient across both major platforms. I do have an Android, so I talk to Gemini on my phone as well, but Claude is my primary work driver.
Kaitlyn Kirkaldy (26:00) Let’s talk about PMs submitting Pull Requests (PRs). Some companies still struggle to give PMs read-only repo access. How did Babylist transition to PMs actively pushing code?
Emily Levada (26:39)
In February, our CEO explicitly told the entire tech team: “Stop what you’re doing. I relieve you of your roadmaps. The most valuable thing you can do right now is figure out a new way of working.”
We took an eight-week window focused entirely on skill-building and new workflows. It wasn’t just PMs and designers—our site merchandisers and marketers started pushing PRs too. For example, a marketer recently edited the codebase to pull a registry photo data source into Iterable for an email campaign.
Because buy-in came directly from our CEO and CTO, we didn’t have to fight for access. Is there still a learning curve? Absolutely. Sometimes a PM builds a prototype with Claude, and an engineer points out it isn’t architecturally sound. But PMs have always filled organizational gaps—historically in design, analytics, or business—and now AI gives non-technical PMs the tooling to fill engineering gaps when needed.
As engineering speed increases, the true bottleneck shifts to knowing what to build. We teach PMs to be full-stack while building product sense within our engineering org.
The biggest win is loop closure speed. Recently, a PM had an idea, realized the necessary data wasn’t instrumented, instrumented it herself in the codebase, waited two weeks, pulled the Snowflake data, and validated the opportunity—all without distracting an analyst or engineer.
Kaitlyn Kirkaldy (31:21) Closing the loop without interrupting engineering leads for basic answers is massive.
Emily Levada (31:37) It transforms information availability. We have a 15-year-old codebase, and I once received conflicting documentation about a legacy feature. Once I had repo access, I asked Claude to map the codebase and generate complete documentation across platforms. It’s like developing an entirely new sense for product managers.
Adam Piel (32:31) It eliminates being blocked. If you don’t know how to use an AI tool, you can literally ask the AI to teach you. It never judges a question as stupid. It’s incredible for onboarding, asking about obscure internal acronyms, or digesting heavy docs.
Emily Levada (34:51) You have to make AI work for you, not against you. My favorite resource for this is Hillary Gridley’s Couch to AI 5K.
I keep local directories filled with meeting transcripts, memory files, and project contexts so Claude has complete visibility into my world. I even built a custom “weekly reflection” skill. Once a week, I ask Claude to audit my performance against known personal failure modes—like failing to block calendar time for high-priority items or blocking team decisions for over 24 hours.
I also built custom writing skills so Claude matches my natural voice. Claude defaults to heavy “AI-speak,” so I regularly instruct it to strip out corporate AI jargon. I also dictate via WhisperFlow and instruct Claude to retain my exact vocabulary rather than paraphrasing. The more effort you put into training your setup, the more output you get.
Adam Piel (37:46) I love the concept of an AI auditing your performance weekly—I am definitely stealing that!
Kaitlyn Kirkaldy (38:22) What do you see as the single biggest challenge or trap for product managers in an AI-mature environment?
Emily Levada (38:59) The core constraint has moved from build capacity to judgment capacity. Judgment doesn’t automatically scale by adding tools.
While tools can pressure-test your thinking, unmediated customer understanding, deep domain context, and product judgment remain the scarce resources. Team sizes and compositions may shift because build capacity is no longer the bottleneck, but PMs must remain fiercely focused on knowing the right things to build.
Kaitlyn Kirkaldy (40:32) You cannot outsource product sense or decision-making. AI gives you tools to make better decisions, but it can’t make them for you.
Lightning Round
Adam Piel (41:01) What is a memory of a product or initiative you shipped that failed spectacularly, and what did you learn from it?
Emily Levada (41:43) I had to recall a physical product once, which was super painful!
On the software side, I spent years at Wayfair building a new Content Management System (CMS). We built a powerful tool tailored for our power users. But as the company scaled, our merchandising team grew from 10 to 60 people in four months.
The system was pushed onto users and use cases it was never designed for. Two years later, it became the system everyone loved to hate. It taught me two critical lessons:
- Never over-engineer or over-generalize a tool beyond its intended use case.
- Respect the product lifecycle, and be willing to sunset things you worked hard on when they no longer serve the organization.
Adam Piel (44:51) Earlier you mentioned specializing in a specific user type (parents/families). What are your broader thoughts on being a specialist vs. a generalist in product management?
Emily Levada (45:22) I fundamentally view myself as a generalist and love exploring new industries. However, in today’s candidate-heavy hiring market, positioning yourself as a pure generalist is tough. When companies have hundreds of applicants, they often default to the person with specific domain experience. Being able to articulate a clear, specific narrative about why this domain is the one you want makes a massive difference.
Adam Piel (46:49) Do you have any hot takes on working in tech in the Boston area compared to the West Coast or Austin?
Emily Levada (47:09) MIT President Sally Kornblith framed it brilliantly: Boston’s bet is “AI + Something.” It’s AI applied to real-world domains—healthcare, biotech, robotics, energy, and commerce. Silicon Valley builds the foundational models; Boston builds what you actually do with them.
Boston has decades of applied ML lineage—from MIT Media Lab and Nuance (which powered Siri) to Alexa’s speech team, Spotify’s personalization office, Wayfair, and Whoop. At Babylist, we have 15 years of proprietary family data. Applying AI to unique, defensible datasets is what Boston has always done best.
Kaitlyn Kirkaldy (48:53) Thank you so much, Emily, for sitting down with us! This was a fantastic perspective on building product teams, using AI thoughtfully, and why Boston is a sneaky AI powerhouse.
Emily Levada (49:14) Thank you, it was so much fun!
Adam Piel (49:14) Thank you so much, Emily!
