Julia Stiglitz: Most Companies Aren’t Even Close to AI-Native
Julia Stiglitz, co-founder and former CEO of Uplimit, joins Greg Shove to explain how her company pivoted twice in the AI era — ending in an acquisition by Handshake — and what she learned building an AI-native team of 30 people along the way.
Uplimit started in 2021 with a thesis that AI could make online learning actually work. A year later, ChatGPT launched and forced a full rethink. Julia and her team discovered that the AI-powered learning platform they were building mattered more than the content on it — eventually arriving at the “AI Skills Instructor,” a system that generates personalized coaching experiences instead of static courses. That product became the foundation for Handshake’s new education vertical, built for a platform with 1,600 universities, 25 million students and alumni, and a million employers.
In this episode, Julia and Greg discuss:
- Why Uplimit pivoted away from content creation entirely, and how hard it is to let go of a product you’ve already invested time and energy building
- The “work slop” problem — when AI makes individual output faster but pushes editing and review burden upstream to leadership
- Redesigning product workflows so designers prototype in code instead of Figma, and what that does to team dynamics
- Why workflow and role changes are harder to drive than individual AI adoption, especially in larger organizations
- The “barbell” hiring model — pairing AI-native early-career hires with experienced judgment — and why she believes the “AI is killing entry-level jobs” narrative is overstated
- Why discipline and focused vision matter more than ever now that building and discarding prototypes costs almost nothing
- Why enduring skills — agency, ownership, curiosity — will separate the top 1% of workers from everyone AI brings to “upper middle”
If you’re navigating a pivot, rethinking how your team builds, or trying to figure out whether AI adoption is changing your org for better or worse, this conversation is a candid look at what that process actually requires.
Transcript
I think in this environment, I actually think is one of the more
Julia:important things if you wanna build a really big company because the
Julia:technology is changing so quickly, and the insights that you have after you've
Julia:gotten started might be dramatically different than the insights you
Julia:had in the very beginning, and that might lead you to, need to let go.
Greg:Welcome to "Super Companies."
Greg:This is a podcast where I interview founders, builders, operators
Greg:around how to build the most
Greg:valuable companies in the world.
Greg:I think those are super
Greg:companies, meaning I think they'll be the most efficient, fastest-growing, and,
Greg:uh, valuable companies in the coming next 10 years as we migrate from a knowledge
Greg:economy to an
Greg:intelligence economy or from knowledge work to intelligence work.
Greg:so welcome.
Julia:Thank you.
Julia:Thanks for having me
Greg:so Julia, let's talk about Uplimit first,
Greg:the company that you were the co-founder of and CEO of,
Greg:and especially the last couple
Greg:years in terms of what you were seeing, and
Greg:then we can talk about, how AI changed the way you built
Greg:products at Uplimit,
Greg:how it changed how you
Greg:worked internally,
Julia:Yeah, sure.
Julia:So, we started Uplimit in 2021.
Julia:we had this thesis that, we could make online learning actually work, and
Julia:the core of that thesis is that we were gonna be able to use AI to do so.
Julia:and so this was a year before ChatGPT launched, but AI was
Julia:always part of our thesis.
Julia:My co-founder, Sohrab Bahaj, had been an early Coursera engineer, and
Julia:then had gone on to Google Brain.
Julia:and so I remember in our pitch deck, which was, pre-product when we raised our
Julia:seed round, and it was, you know, a very loose vision of how AI would translate
Julia:into, e-efficacy in terms of the product.
Greg:And what was the primary thesis there?
Greg:Was it
Greg:basically personalization that AI would
Greg:make the learning experience?
Greg:and I think by default we think that all
Greg:experiences will be more
Greg:personalized, right?
Greg:and more relevant.
Greg:Is that sort of your
Julia:yes, and I would say it was particularly around the
Julia:support aspects of learning.
Julia:so how do you support learners to successfully complete?
Julia:how do you nudge them?
Julia:How do you follow up with them at the right time?
Julia:and, then how do you give them the right information or, the right content, so
Julia:that they can unlock their learning?
Julia:it was a lot around that, that we could, you know…
Julia:y- It's really expensive to use humans.
Julia:If we could use AI, we could provide the support that you
Julia:would get from more a small group instruction or a one-on-one, teacher.
Julia:And so that was really the thesis.
Julia:and we, saw that we could drive engagement, we could drive completion
Julia:with those sort of supports in place.
Julia:and then a year in, ChatGPT launched, which, you know, I, I don't think
Julia:anyone, anticipated, and it really made us, s- like rethink our whole roadmap,
Julia:and what we were building and what the opportunity was and how we could,
Julia:go to market
Greg:So making education work, obviously what everybody
Greg:needs to be trained on or educated on
Greg:is AI s- in the
Julia:Yes.
Julia:Yes.
Greg:AI skills
Greg:and AI capabilities.
Greg:So what you were
Greg:seeing that was working or not working
Greg:as it relates to AI upskilling?
Julia:so a couple things happened, about that time for us.
Julia:we started selling into companies, so working with large companies who
Julia:wanted to upskill their workforce.
Julia:we actually had a whole catalog of, courses that were focused on AI training.
Julia:we had partnered with really incredible, engineers, who were practitioners and
Julia:built out these, very cool courses, teaching people how to, use AI.
Julia:it was technical courses, So these were the, engineering
Julia:courses, focused on engineers.
Julia:and, we went to companies who wanted to use them to help their
Julia:engineers, um, learn these skills.
Julia:on the content front, I think we were too early.
Julia:I think companies, were just starting to talk
Julia:about upskilling and, reskilling AI,
Greg:the curse of every entrepreneur,
Greg:right?
Greg:being a little early and sometimes a lot
Julia:Yes.
Julia:And at the same time, where we saw, the buyers' eyes light up was
Julia:with some of the AI tools that we had built to support the learning.
Julia:so like AI content creation tools, that could, generate content?
Julia:a tool where you could, write your answer in and get feedback, AI role plays as we
Julia:sort of kept on going down this direction.
Julia:And quickly we saw that the platform we were building was actually much
Julia:more valuable than the content?
Julia:at that time, and that there was just so much room to sort of reimagine what
Julia:an education platform looked like with this technology that could do these
Julia:things that when I was at Coursera we could only dream of And so we just
Julia:sort of went headfirst into building out an AI-first learning platform that
Julia:could support corporations, in their upskilling, which included AI upskilling,
Julia:but was also much broader than that.
Julia:it was, leadership and customer education and sales and all the different areas
Julia:that companies upskill their workforce on
Greg:and did that pivot start to work?
Julia:Yes.
Julia:it did.
Julia:we brought on some, um, really amazing customers.
Julia:Databricks started using us to, scale their customer education program.
Julia:Uh, one of the top consulting firms in the world, uh, con- continues to use us
Julia:to, train all of their new consultants.
Julia:We work with Kraft Heinz and GE Healthcare
Greg:and so what happened after that?
Julia:building and, you know, it was a very interesting time to build because the
Julia:technology was shifting under our feet.
Julia:one of the things that I learned in building in this environment
Julia:is you just can't be too precious about what you're building.
Julia:Like, you have to be very willing to throw it out.
Julia:we saw that with our AI authoring tool where we, built the first version of it.
Julia:It was okay, and then the technology changed and we need, you know,
Julia:essentially needed to throw it out.
Julia:And at some point we said, we were building ways to make content easier to
Julia:generate, and then, we asked the question, like, "Do you need to generate content?"
Julia:even that was so fundamental to the job of L&D, creation of
Julia:courses, pushing out courses.
Julia:do you actually need to do that?
Julia:And instead could we create a system where you don't create content?
Julia:The L&D person puts in context, and it generates, for each individual,
Julia:a personalized learning experience that they get to go through.
Julia:So there's no content getting produced that is edited and, sent to everybody.
Julia:And so that was called the AI Skills Instructor, which is, what we have today.
Julia:it's a really awesome way for companies to train people on whatever skills that
Julia:they want, and have each person have their own individualized experience
Greg:So it's not just AI
Greg:skills.
Julia:It's not just AI skills.
Julia:It's actually for companies it was mostly, the, on the uplift
Julia:side, is mostly leadership skills.
Greg:I was gonna say it's is that the thing that
Greg:Handshake wanted
Greg:Is, is that what they saw as, as
Greg:significant potential for them?
Julia:I had known, Gareth, uh, and John, um, the leaders of
Julia:Handshake, for about 10 years.
Julia:Um, and, um, we started talking earlier in the year, and they had
Julia:this vision around building out an education vertical, on Handshake.
Julia:And they'd had this vision for a long time, where there's 1,600 universities
Julia:on the platform, 25 million students and alumni, and you have a million employers.
Julia:And, there's gaps that they saw in terms of especially young people as they're
Julia:trying to get jobs right now in the economy, and that there's an opportunity
Julia:for education and education vertical to help bridge some of those gaps.
Julia:and this is a time where, hiring for young college students is down.
Julia:It's about 7% year over
Julia:year by some estimates.
Julia:in tech companies, it's, like, 25%, down.
Julia:Um, and some of the skills that they're seeing that there's a
Julia:real need for are AI skills.
Julia:and so we started talking about that and, realizing that, there…
Julia:I think that there's, right now, this sort of arbitrage opportunity
Julia:for, young people and, you know, any professional to learn a set of AI
Julia:skills that can be really differentiated as they're applying for jobs and as
Julia:they're, showing up in the workplace.
Julia:And it's not, months of work or years of work, to get these skills.
Julia:You actually can make a lot of progress, in a week.
Greg:So your last product
Greg:at Uplimit, AI skills instructor,
Greg:AI-powered learning or co- I would call
Greg:it coaching.
Greg:It's not really learning.
Greg:You're right.
Greg:You know, you're not
Greg:pushing content to people.
Greg:You're basically generating these interactive real-time,
Greg:coaching experiences.
Greg:and initially it was for
Greg:any skills.
Greg:post-acquisition by
Greg:Handshake, clearly AI
Greg:skills are in huge demand, and, I think a lot of employees
Greg:and employers are struggling with that, and I wanna talk to you
Greg:about that, what you've seen in terms of why aren't we seeing this
Greg:upskilling happening faster, and why the, workforce transformation
Greg:is not happening faster.
Greg:But Handshake's got a big business currently
Greg:with universities, with students, like that whole ecosystem, I
Greg:guess.
Greg:Plug in
Greg:this new capability to generate personalized learning
Greg:that, is AI powered and it's also teaching you AI skills for the
Greg:individual at scale
Greg:very cost-effectively.
Julia:that's right.
Julia:And we,
Julia:uh, launched to consumers.
Julia:Uh, we'll be going to enterprise soon, and then we did it in partnership with the
Julia:leading AI companies, so OpenAI, Google,
Julia:Vercel, Clay, Lovable, Figma, a whole set of companies that
Julia:we developed the courses with.
Greg:Got it.
Julia:Yeah
Greg:are they courses still,
Julia:we're calling them missions.
Julia:and so they are not courses, in any way other, in terms
Julia:of how you would think of it.
Julia:They're, 10 to 40-minute experiences where you build something.
Julia:you build an artifact, and
Julia:you post that artifact on your LinkedIn profile or Handshake profile.
Julia:and it's intended to get you a job, and it's not credentialed.
Julia:We don't have a huge skill taxonomy associated with it.
Julia:it's really about building and showcasing what you can build and, being able to talk
Julia:about what you've built And share that with employers and apply those skills back
Julia:to projects that you're doing on the job
Greg:I'm just curious, the B2C model, does the consumer pay?
Greg:Or is it it's free?
Julia:It's free.
Julia:At this point it's free, and then there will always be for our students it'll
Julia:be free, and then, for professionals most of it will be free, and then for
Julia:companies it'll be a freemium model
Greg:Okay, so I do wanna now talk about your last couple years at
Greg:Uplimit as CEO, meaning you're now, you're pivoting the product a lot it
Greg:seems like, and yet
Greg:you made a great observation, I think we all feel it right
Greg:now, which is the rate of
Greg:change.
Greg:And so you, you might have made product roadmap
Greg:decisions and/or built products, and you gotta basically throw them away,
Greg:abandon them, correct course, whatever.
Greg:when do you do that too much or too little is one of my questions.
Greg:But I want you to talk
Greg:first about how were you trying to change the way Uplimit
Greg:was working for the last
Greg:couple years?
Greg:were you trying to become AI at least augmented?
Greg:in terms of everybody, not just the product team or the software team, but
Greg:finance, HR, marketing, sales, whatever.
Greg:Uh, and your own personal,
Greg:and your own personal work as CEO.
Greg:and how did you think about how demanding you were of your own team around becoming
Greg:AI, AI augmented or even AI native?
Greg:Cause you and I were, both started these companies prior
Greg:to this generation of technology and especially the last 18 months in terms of
Greg:Claude Code, Cowork, and now of course
Greg:Codex GPT.
Greg:So just talk about that in terms of how you saw how you guys
Greg:were changing the way you worked, trying to accelerate everything, uh, work more,
Greg:work more efficiently, and how hard that
Greg:was,
Julia:Yeah.
Julia:So a lot of the change that we saw happened organically.
Julia:and we were lucky, or w- I wouldn't say lucky, that the conditions of an
Julia:early stage startup that are real- where you're really small, where you have a
Julia:set of people that are risk-taking early
Julia:adopters, are, like, uh, where you don't have v- very much of any sort of
Julia:organizational process in the beginning,
Greg:And everybody's overworked, right, as
Julia:Everybody's overworked, you're resource constrained, you're,
Julia:like, looking for any opportunity you can get to, um, do more.
Julia:you don't have all these subject matter experts that are
Julia:protective over their domains.
Julia:So, it lends itself well to, early adoption, of these technologies.
Julia:So the engineering team started with Cursor, in about a year,
Julia:January of last year or so.
Julia:they moved over to Cloud Code.
Julia:and then across the board, our salespeople were using, OpenAI and summarizing all of
Julia:their, meetings that they had and sending out those notes and, ended up using Cloud
Julia:Design for a lot of the pr- presentations
Greg:how big was the
Greg:team at
Julia:30 people.
Greg:were you modeling that
Greg:behavior?
Greg:Like how much of this came from you or did it
Greg:come, you said it maybe more organically, And what were your, what were your
Greg:expectations of the team in this respect?
Greg:Like how
Greg:much did you want to
Greg:see them relying on AI?
Julia:we definitely pushed the
Julia:team to use it.
Julia:we gave them access to the tools.
Julia:We encouraged them to use it.
Julia:We set up a Slack channel where people were showcasing different use cases.
Julia:in our demo day The first section of our demo day, we would discuss
Julia:different things that, people are doing with AI and what they had built.
Julia:there was a lot of people experimenting, both with things that they were
Julia:doing at Uplimit, but also on their, own individual side projects.
Julia:that they were also showcasing, what they built.
Julia:one of my co-founders, Jake Samuelson, is just an amazing tinkerer.
Julia:I would say h- in some ways, he was an individual force behind this,
Julia:where he was playing all the time and showcasing all the time what he
Julia:was building and what was possible.
Julia:but we had a lot of other, leaders who just w- were playing and showing, and I
Julia:think that really changed the culture.
Julia:One of our designers, both of our designers actually, just went deep
Julia:into prototyping with, AI and starting to think through new workflows for
Julia:how the engineering team and product team could work, which I can…
Julia:and share about.
Julia:but that was…
Julia:There-- So there were these examples across the company of
Julia:people doing really cool things.
Julia:And on go-to-market, it took Like, a more of a push, but
Julia:not very much of a push, just
Julia:given, how early we were.
Julia:so that's what we saw.
Julia:But once people started using it, there was sort of a couple
Julia:of challenges that we saw that
Greg:yeah, say m- say
Greg:more
Julia:Okay.
Julia:So one of the challenges that I think a lot of people are seeing right now is,
Julia:yes, you want everybody to use it, but you want them to use it in ways that are
Julia:productive, that cause organizational efficiency and not inefficiency.
Julia:And, we had hired…
Julia:As an example, we had hired a new marketer, and all of a sudden, I
Julia:found myself Editing heavily every single blog post that was coming out,
Greg:Work slo- work slop.
Greg:Work
Julia:Y-
Julia:yes, and it was this sort of, transference … Yes,
Julia:it was m- it was so much more efficient for this marketing person.
Julia:Like, this marketing person could produce blog posts in 30 seconds,
Julia:and the work just got pushed upstream
Julia:to me.
Julia:and then, you know, I was editing everything.
Julia:And so, we actually had to put some constraints on how to think …yeah,
Julia:actually you, you…
Julia:first of all, certain people we were like, "Okay, you need to pause your
Julia:AI writing, and you need to sort of write without AI for a little while."
Julia:and, we sort of, pulled it back.
Julia:and I think that's…
Julia:one of the larger changes that I'm seeing right now, as people are pushing
Julia:AI adoption that right now it's gone beyond just, like, use AI, but really you
Julia:need to own the work that's coming out.
Julia:and so people are much more thoughtful about that but that was definitely
Julia:one of the challenges that we started seeing where people were just
Julia:shipping things with AI, not, you know, not reviewing it, which causes
Julia:real organizational inefficiency.
Julia:and it causes inefficiency at sort of higher levels, which
Julia:is even more problematic.
Greg:My, my favorite story of that is getting
Greg:the internal
Greg:business case to approve a $500 decision, and it's like three
Greg:pages and beautifully formatted.
Greg:And it's like, first of
Greg:all, why the fuck am I reading this?
Greg:And why have you sent me a business case for a
Greg:$500 decision?
Greg:Just go spend the $500.
Greg:Like, what's going on here?
Greg:But just because you can, doesn't mean you should.
Greg:And you're right, it gets pushed up.
Greg:Now it's friction for me to deal with,
Greg:Anyway, what else did you see or anything
Julia:Yeah.
Julia:So one of the other changes that I would say was harder, took a little
Julia:bit longer and and this is I think one of the places where it, is even
Julia:m- more true in larger organizations is it's easy to change individual
Julia:productivity, or relatively easy.
Julia:you get people to use these tools, they get the aha moment, they get
Julia:a couple workflows that really work for them, it makes their
Julia:life easier, And they will use it.
Julia:And you need to get them over that hump of like, "I'm intimidated
Julia:by this technology," or, "I don't understand the use cases."
Julia:But once you do, then they use it and it's great.
Julia:What's harder are things that, where you have to change workflows
Julia:and team dynamics and things that touch people's identity more.
Julia:and so if you think about, product and engineering, which is the place where
Julia:people's roles have changed the most.
Julia:You know, what an engineer does in a given day is totally different today
Julia:than what it was, two years ago.
Greg:Months ago
Julia:Six months ago.
Julia:Yes.
Julia:Yeah.
Julia:And, you know, the designers, the product, managers, they can build with
Julia:these tools, and the engineers, they can design and they can think about products.
Julia:And so there is a real blurring of some of those roles, in
Julia:ways that need to be rethought.
Julia:And so, one of our designers has put forth this sort of new process where instead
Julia:of building just in Figma, the design and the product team actually build full
Julia:prototypes of the entire experience.
Julia:Um, so right now with what we just launched, we have an entire
Julia:prototype in Vercel of the same thing that is live, uh, in production
Julia:on Handshake.
Julia:and when we wanna make changes, the designers go in there and they
Julia:build the changes in the prototype.
Julia:They really understand what the interactions look like, how it works,
Julia:and then it gets moved to engineering, and then engineering builds from there
Julia:but that's a really different flow.
Julia:and it's a really different experience.
Julia:and as an executive, it's a different experience too 'cause you, you know, very
Julia:quickly can see what it's gonna be like.
Julia:Like, it's, it's not just imagining from a wire frame.
Julia:You can actually play around with it.
Julia:and You don't need t- to debate certain things.
Julia:you can have people just go off and try to build it, and then You have to go through
Julia:certain processes to actually get into production to make sure it's secure and
Julia:it's scalable and all of those things.
Julia:But, that's a different flow, and I would say it took us a little while to get
Julia:there, and it, there was definitely some head butting across the team as we were
Julia:trying to get there, where, initially our designers were trying to ship code
Julia:that would go to production, and the engineers were like, "What is this?
Julia:This is, like, 20,000 lines of code," and you could do it
Julia:in a, you know, 1,000 lines.
Julia:And, like, that, again, that is not efficient.
Julia:Like, that is inefficient, where you are creating work for another individual
Julia:to then have to review that PR.
Greg:Yeah, no, I totally get you.
Greg:I was asked by the head
Greg:of, AI
Greg:transformation at a pretty large, legacy firm, technology firm.
Greg:Uh, he asked
Greg:me, you know, "What do I need here to
Greg:really make this transformation work across, you?
Greg:know, global large workforce?"
Greg:let's say 100,000 people.
Greg:my answer was,
Greg:you know, in jest but
Greg:serious, was, You probably need 1,000
Greg:or more," I call them workflow redesigners, people call
Greg:them FTEs,
Greg:or, you know, who is actually in there rewiring these
Greg:workflows, and 1,000 therapists."
Greg:And we kind of laughed about it, but he's like, "No, I get it, like the…"
Greg:And he's like, "The therapists."
Greg:"Yeah, like, well, the therapists are gonna have to go in there and
Greg:really, you know, sit and hold hands with everybody and sort of
Greg:c-coach them through this, and that your job did change overnight, and that
Greg:y- you know, this is how you will be
Greg:valued.
Greg:But don't worry, you will be valued or not," or, you know, or, "Your job is gone.
Greg:I need to help
Greg:you go figure out what to do next," or whatever it.
Greg:might be.
Greg:But, like, there is a real h- obviously human component to this.
Greg:We all know
Greg:that by now.
Greg:I think it's obviously,
Greg:in large
Greg:organizations, just, a, a big challenge, obviously.
Greg:the numbers are bigger, and I think people are a little more stuck in place,
Greg:mentally and organizationally.
Greg:The boxes are, thicker on the org chart, and so
Greg:this kind of change is tougher.
Greg:wouldn't it be easier to start a company
Greg:fresh today that is AI native from day one?
Greg:'Cause you don't have the boxes on the org chart.
Greg:You don't have all these legacy
Greg:systems.
Greg:where the hell is that data that we need Claude to
Greg:go get?
Greg:Like, how do you think about that?
Greg:I'm
Greg:just curious.
Greg:do you think about, "Oh, it'd be great…"
Greg:I'm not saying you, uh, not necessarily tomorrow.
Greg:You sold the company.
Greg:You should hang out at
Greg:Handshake for a while and, and, you know, focus on other stuff.
Greg:I don't know, your kids
Greg:or, or, you know, whatever hobbies you
Greg:have but at
Greg:some point, d-do you think about, okay, would, it would,
Greg:be really different to start a new company now and how it might be different?
Julia:I think, I mean, across many dimensions, Like, I felt this at Uplimit.
Julia:letting go of a product, is really hard, and that's at a small company
Julia:where you don't have a huge amount of users, but you have invested time
Julia:and energy in that e- effort, and maybe you have some customers in it.
Julia:so moving off something to go to, the thing that you think could be bigger
Julia:but is a lot riskier, is really hard.
Julia:So on the product side, I would say, just on the product strategy side, there were
Julia:many times over the course of Uplimit where we had this conversation of like,
Julia:do we just completely, pivot from the product that all of our customers are on
Julia:to go to something that is more aligned with what the technology is now capable
Julia:of and where we believe it's going.
Julia:and it is…
Greg:how would that be different now?
Greg:But if you were
Greg:starting a new company today as a, founder,
Greg:CEO, operator, like how would that be different?
Greg:Do you think meaning you just, you would have even less ownership
Julia:it's … Yeah, so on the product strategy side, I think
Julia:you get to start from scratch.
Julia:Like, you can, and you know, technology will keep moving, but I think there's
Julia:a better understanding of what the technology can do and where it's going
Julia:and, some of the opportunities there.
Julia:I think that it is still an environment where everything is
Julia:moving, so I I actually think it's gonna continue to change.
Julia:but I think you do get to start from scratch, from the
Julia:product strategy perspective.
Julia:So I think there's that.
Julia:On the team perspective, I think it does shift how you hire
Julia:and how many people you hire.
Julia:what it looks like, especially in the early days when you're proving
Julia:it out, you don't need as many engineers to build something and
Julia:see if people want it, which is the hardest thing in building a startup.
Julia:Like, does anyone
Julia:actually care about this?
Julia:yeah,
Julia:And wanna pay for it?
Julia:And so I think you can do that in a much more cost-effective way.
Julia:You can sort of go through, "Here are the biggest risks and hypotheses that we
Julia:have," and figure out if that's valid or not valid, and sort of work through it.
Julia:So I think there is a very different way of building.
Julia:that … And the, and the new capital implications.
Julia:we raised pre-product.
Julia:and we needed to raise pre-product.
Julia:Like, we were paying our designer, who was the one sort of employee that we had
Julia:before we raised from our credit cards.
Julia:And it was like, we needed to raise.
Julia:and we wouldn't have been able to move how we moved without it.
Julia:That's not true anymore.
Greg:You know, one of the things I worry about, well, several things.
Greg:well, there's lots I worry about.
Greg:But in terms of this aspect of, starting or running startups,
Greg:a couple thoughts come to mind.
Greg:I'd be curious what you think.
Greg:For someone like me who's, and this is the way I describe it, my
Greg:mi- my wife might say I have ADHD.
Greg:I describe it as I'm in- I'm intensely curious and always, attracted to new
Greg:opportunities and
Greg:perhaps
Greg:moving on too
Greg:fast from what we were already doing.
Greg:And of course,
Greg:you can move on fast in these moments 'cause you can prototype that much faster.
Greg:You do get market feedback faster.
Greg:Which leads to my second concern is maybe we're not
Greg:sticking enough with the harder problems
Greg:that
Greg:just do take longer to figure out.
Greg:And if you look at great companies, what we used to call generational
Greg:companies, now I call them super companies, sometimes yeah, you gotta
Greg:like work really fucking hard for a long time and not get great
Greg:market signals, uh, and eventually, you know, kinda dial in
Greg:that product market fit into a much bigger and significant
Greg:opportunity and then company.
Greg:and maybe we won't
Greg:be doing that as much 'cause we'll, we'll, we'll be able to move on
Greg:fast and give up because…
Greg:and by the way, we haven't made as significant investments 'cause we've
Greg:just been prototyping and vibe coding, so throwing it away is not a big deal.
Greg:I'm just curious how you think about, this going forward
Greg:as an
Greg:entrepreneur.
Julia:it definitely resonates.
Julia:I remember in the early days of Coursera, there was so many
Julia:ideas that were tossed around.
Julia:in retrospect, it was probably good that we couldn't execute on all of
Julia:them, and instead stayed focused on Coursera's, core competitive advantage,
Julia:which was the university partnerships.
Julia:it sort of resonates that, if you could build faster, would you?
Julia:but I think it's like, what does it mean to lead in this environment?
Julia:That sort of discipline, focused vision, is what's required of
Julia:leadership in this environment.
Julia:So it's like, yes, you can build all these different things, but you need to think
Julia:through what the implications of it are.
Julia:I was having a conversation with a founder the other day, and he wanted
Julia:to go after all these different customer segments at the same time.
Julia:And, you know, technically they could build a lot of
Julia:these different slight variations of the product that would allow them to
Julia:go after a bunch of these segments.
Julia:but, you know, what are the implications in terms of the brand?
Julia:And what are the implications in terms of how you go to market and
Julia:the channels that you start building and, how people associate with you?
Julia:And then longer term, like how that sort of shifts the product roadmap in
Julia:ways that you want or not, don't want.
Julia:So I think it's, I see it more as it's, a requirement of leadership, and it's
Julia:always been a requirement of leadership, but now you just have more tools in your
Julia:disposal that makes it even more important
Greg:Yeah, I was gonna say it just seems to
Greg:amplify, this kind of ability and this willingness to kind of you
Greg:know, the balance, I guess, of being decisive and ambitious and
Greg:aggressive, but also knowing when to just kind of slow down and be
Greg:patient and sort of double down and, stick with your, either
Greg:your intuition or maybe what the data says.
Greg:Hopefully, there's enough data
Greg:at some point emerging from any of your efforts, right?
Greg:That you can say, "Okay, one's working,
Greg:this one's not."
Greg:Yeah, I, I do think, I do think we'll be flooded with shitty
Greg:software in the coming years
Greg:just because it's so easy to build software.
Greg:a lot of it won't be good and will therefore probably not
Greg:find any product market fit.
Greg:Julia, I know you've only been at Handshake a few months, and they are,
Greg:of course, not, AI native either in terms of what we were talking about.
Greg:I'm just curious if you've
Greg:seen anything interesting at Handshake in terms of
Greg:how they work, in terms of relying on AI that you can share that maybe we can
Greg:learn from, and if not, I've got a bunch
Greg:of other questions, so don't worry.
Julia:So, I mean, I would say a couple things.
Julia:Like, one, I would say all the execs are pretty fluent in AI.
Julia:You know, they are building their own decks, with it.
Julia:and they are, you know, they have their daily digest coming from um,
Julia:OpenAI and, um, uh, are, themselves playing with it, regularly And sort
Julia:of showing what they're building.
Greg:which
Greg:I would, which I'd argue is table stakes at this point, right?
Greg:I mean, if you're leading a
Greg:company, certainly a startup, but any, I
Greg:think almost any company, But it feels like
Greg:if you're leading a company of at least zero to 1,000 employees in anything
Greg:related to tech or tech data services,
Greg:whatever, if you're not a
Greg:hands-on keyboard, then I think you should go work on your golf
Greg:game, basically.
Julia:yeah.
Julia:And then I would say the engineers, you know, they're all using Cloud
Julia:Code, and that's how they're building.
Julia:and then there's a lot of freedom and encouragement to rethink process.
Julia:So the process that I mentioned that, we were doing at…
Julia:starting to do at Uplimit in terms of prototyping and designers prototyping and
Julia:then, having a mirror of the site, it- we've been able to bring that over here.
Julia:and one of our designers is starting to lead that effort internally
Julia:here at Handshake to do something similar, which is really exciting.
Julia:So there's, I would say, Like, appetite, excitement around,
Julia:rethinking workflows with AI.
Julia:and so it's, that, you know, that's not…
Julia:it's not a battle, it's more like, "Oh, that's awesome that you're doing
Julia:that Like, "Let's do that, Let's figure out how we do that across the board."
Greg:I'm just curious, at
Greg:Handshake or, or p- or previously at your startup at Uplimit, did you look at this
Greg:revenue per employee number as a measure
Greg:of efficiency, and did you talk about it?
Greg:And do you guys think about it at Handshake currently, at the executive
Greg:level, which is, you know, how efficient are we, and is AI making
Greg:us more efficient, meaning, you know, more revenue per employee?
Julia:We didn't look at it at an up limit.
Julia:I'm not sure if Handshake is looking at it more broadly at this point.
Julia:they're definitely, watching employee spend with
Julia:it.
Julia:but I'm not sure if it's tied to
Julia:efficiency at
Julia:this
Julia:point.
Greg:okay, I wanna talk about, young people, 'cause you said, offers are down.
Greg:I think, like, you said some, the, some of the jobs data, uh, and particularly
Greg:in tech, is not good for entry level.
Julia:Yes.
Greg:Do you, Do you,
Greg:think that's AI or just a broader economy and still a continued
Greg:contraction of workforces that were too large, post-pandemic?
Greg:Which seems a long time ago now, but I think we're still
Greg:in the era of post-pandemic.
Julia:yeah, so I think it's mostly, broader economic forces.
Julia:especially as I've I meet with and talk to a lot
Julia:of large companies, and the
Julia:level of AI adoption is not at the level at most of these companies
Julia:where I would expect to see, major changes in terms of their core hiring practices.
Julia:But What I don't want is that becomes a narrative, and I think
Julia:it is becoming a narrative,
Julia:that you don't need to hire early career talent because of AI.
Julia:and in fact, I think, what I would like to see is sort of the opposite to be true.
Julia:Like, this is, you know, young people who have not been in the workforce
Julia:do not have a predisposition towards certain workflows, do not have a
Julia:certain identity attached to, being an engineer in a certain way, are gonna
Julia:be some of the most malleable early adopters, enthusiasts of a lot of
Julia:the transformation that is happening.
Julia:And I think that there's a massive opportunity to, to leverage
Julia:them, to help transform your organization if done thoughtfully.
Julia:And so, at our, the other day at the we had a big summit.
Julia:One of the people that spoke was the chief people officer of Notion, and she talked
Julia:about this barbell approach that they're taking to hiring, where they're hiring
Julia:a lot of early college grads and young people, and then they have more senior
Julia:experts, and that they are, intentionally matching them together, where, the young
Julia:people that they're hiring are, you know, enthusiastic and, and, AI native, and,
Julia:and, and the more experienced people bring judgment and, and, and, and whatnot.
Greg:do you buy that model?
Greg:would you do that in your next startup?
Julia:I would hire, you know, AI native talent that comes in that has…
Julia:is able to, not think so narrowly about what a function is, and has access
Julia:to these tools that give them outside expertise and allows us to push the
Julia:bounds on what we can do, for sure.
Julia:Like, I think that there's a ton of potential in
Greg:Yeah, I agree.
Greg:when I'm asked by leaders, and everybody wants, like, "What are the things
Greg:I gotta do to make this happen?"
Greg:Uh, and, and I think there are some moves you can make that can sound very
Greg:tactical, but I think they're very impactful, and that's one that's always
Greg:in my top three, which is if you've got one open position in your team,
Greg:does…
Greg:it could be, you could be the CMO and so you're gonna have
Greg:more than one open position.
Greg:But in all these open positions, backfill with a 24-year-old or 23, whatever.
Greg:backfill with someone who's, to your point, really early career.
Greg:They will be AI native, and if they're not, they will in a few months.
Greg:They're gonna be AI native.
Greg:And you're right.
Greg:they're not gonna arrive with any baggage expectations.
Greg:I describe it as what's the cognitive delay between the job they're given, the
Greg:task, the work they have to do, and their decision to go work with AI on that task?
Greg:And they have no cognitive delay, meaning they get the task, whatever
Greg:it is, and they just go right away to GPT or Claude or whatever it
Greg:doesn't really matter which AI.
Greg:And that's what you want, I think.
Greg:you gotta begin to model that behavior.
Greg:So yeah, I, I…
Greg:My point of view is if you got an open req, try.
Greg:I mean, sometimes you can't.
Greg:You need to hire experience.
Greg:I get that.
Greg:but I just think that you're right.
Greg:I think AI is amazing for young people.
Greg:I think AI is gonna be tough on jobs.
Greg:We'll see, um, in the coming couple, three years.
Greg:I think it's gonna be hard on people in the middle, hardest on
Greg:people in the middle, frankly.
Greg:I think it's great for young people.
Greg:I think they can be way more performant, um, and be really, They can, just
Greg:accelerate their economic potential,
Greg:uh, much f- much faster if they, if they work this way, which presumably
Greg:they will 'cause they're AI native at school 'cause they're cheating
Greg:all every day with, with GPT.
Greg:So,
Julia:Um, and I think it can change how companies hire too.
Julia:Um, and so one of the other stories that the Notion CPO shared was that
Julia:they were hiring their class of, uh, BDRs, and what they did to hire this
Julia:class was they put out a challenge for them to essentially build something on
Julia:Notion.
Julia:and then they had people that were using some of the AI tools in Notion
Julia:to build really cool things, and they used the projects to select the
Julia:individuals, who could go through.
Julia:And that's the opportunity.
Julia:it's like the barriers to build or just, do something have, sort of disappeared.
Julia:And so if you are an individual and you take some initiative and you have
Julia:some ideas and you're creative, you can really differentiate yourself,
Greg:I think it's never been easier.
Greg:I really don't think it's been that hard to be in the top 1% of a company.
Greg:It's certainly in the top 5 or 10%.
Greg:I think it's actually a slam dunk, like show up, you know, you know, answer
Greg:your email or Slack on the weekend.
Greg:and now I think almost even easier and/or like you can get the 1% now.
Julia:Yeah,
Greg:at any sort of level in the org, you can be in the 1% of the org I think.
Greg:and it's just not that hard, from my perspective.
Greg:Yeah, it, it takes effort and commitment, and you gotta care and give a shit,
Greg:and not do work slop, care about the quality of your output and so on.
Greg:But I just think it's, again, a great moment to be in the 1%.
Greg:You know, I think you're safe there, very safe there.
Greg:Uh, if you think about your economic sort of, opportunities
Greg:or what risks might be coming.
Greg:So yeah, I want my kids there.
Julia:what's interesting is the differentiators in the 1% actually aren't
Julia:… Like, the AI skills will get you,
Julia:you know, to the middle, or, you know, maybe upper middle.
Julia:But the differentiators in the, 1%, top 10% of employees are gonna go back to
Julia:the enduring skills
Julia:of, agency and ownership.
Julia:Like, you're not just pushing out AI slop, but
Julia:you are really thinking about it and evaluating it and
Julia:retaining ownership of it.
Julia:it's gonna be around your creativity, about your curiosity,
Julia:like, your ability to learn.
Julia:Like, all these things that a good liberal arts education would
Greg:Yeah, I, I,
Greg:no, I think that's right.
Greg:and just give a shit.
Julia:Give it a…
Julia:Yes.
Julia:Yes.
Julia:Care.
Greg:when the boss is texting you at 7:00 PM on a Friday,
Greg:it sucks, but, like, respond.
Julia:Yeah.
Julia:Yeah.
Julia:Yeah.
Julia:Yeah.
Greg:uh, you Showing up matters a lot, and do it consistently.
Greg:you do it once, nice, but do it for a couple years.
Greg:you'll be in the 1%.
Greg:Julia, I'd like you to g- give some advice.
Greg:And, a- and, and if you said, "Don't start an ed tech company,"
Greg:you and I have been there, uh, wor- wor- worthy businesses, hard to exit.
Greg:You exited though, so maybe you're one step ahead of me.
Greg:Uh, and I h- I hope, uh, I hope the, the Handshake 2.0
Greg:or 3.0,
Greg:whatever version of Handshake this is, maybe this is the one that
Greg:goes public or it gets acquired or whatever, and so you, you get to ride
Greg:along on, ride along on that exit.
Greg:But besides, uh, don't start an ed tech company, as you know,
Greg:my daughter works in ed tech.
Greg:she was at Guild, now she's at Go1.
Greg:and so, doing God's work, trying to help others, um, uh, stay professionally sharp.
Greg:it doesn't seem like investors, care as much as we'd like maybe in terms of exits.
Greg:but what other advice do you have for our new founders, operators trying
Greg:to build a super company, trying to build products and services this way,
Greg:and run their companies, AI natively?
Julia:I mean, I do think one of the biggest things on the product side
Julia:is don't be afraid to do a big pivot.
Julia:and that the…
Julia:It is, you, you know, you, you close a couple customers, you have
Julia:a few million dollars in revenue.
Julia:It feels, it feels like you are so far, but in terms of it is th-
Julia:you are at the very beginning.
Julia:and it is so small relative to where you are trying to go.
Julia:And so being willing to let that go, I think in this environment, I actually
Julia:think is one of the more important things if you wanna build a really big company
Julia:because the technology is changing so quickly, and the insights that you have
Julia:after you've gotten started might be dramatically different than the insights
Julia:you had in the very beginning, and that might lead you to, need to let go.
Julia:and trying to do them both for a small company, I think is, is even
Julia:if you're an excellent at execution,
Julia:uh, really hard.
Greg:Yeah, no, listen, that's great advice.
Greg:I often think about this, Julia, because, it used to be…
Greg:And I even, I think that you, there was a bunch of data to back this
Greg:up, that if you got to a million in revenue in a traditional SaaS
Greg:business, you could get to 10, and if you got to 10, you could get to 100.
Greg:the odds just got that much better, as you pass those milestones,
Greg:and that still might be the case.
Greg:but it kind of felt like to me, if I got to 10 million, even three or
Greg:four million, I was like, "Yeah, I got a business, you know, I'm safe.
Greg:I'm not gonna go out of business."
Greg:And m- but, but I think, I think the reality is you're probably not safe until
Greg:you're at 100, and maybe not even 100, i- in terms of million recurring revenue,
Greg:uh, if you're in that kind of business.
Greg:and that to your point, That's where we need to be aiming, and more than
Greg:aiming, like, on a path to get there.
Greg:And if you're at four, three or four or five million, and you can't really
Greg:intellectually, honestly say to yourself you can see 100 then or more, uh, and
Greg:probably a lot more, then you're right.
Greg:it makes sense to me that you gotta call it and either pivot or stop.
Greg:And by the way, I've done so many pivots, I, uh, as I know
Greg:you did at Uplimit, at least two.
Greg:I'm not sure sometimes the right thing is pivoting.
Greg:Sometimes the right thing is stopping.
Greg:But that is, material for our next episode.
Greg:So we'll talk about that.
Greg:I'm doing a podcast later today, meaning I'm the guest, and the
Greg:premise of the podcast is what's your favorite book, as an entrepreneur
Greg:and let's talk about the book.
Greg:And one of my favorite books about entrepreneurship is The
Greg:Hard Thing About Hard Things by,
Greg:by Ben Horowitz.
Greg:And,
Greg:and of course, what's so misleading about that b- it's a great book, and
Greg:every entrepreneur should read it, although maybe not, meaning I think you
Greg:should just talk to Claude about it.
Greg:But, his pivot worked, and that's where the book kind of feeds into the
Greg:mythology of Silicon Valley and the mythology of pivots, which is they work.
Greg:his pivot at Opsware worked.
Greg:Stewart Butterfield's at Slack worked, but many don't, as you know.
Greg:they're much harder to pull off, as you
Greg:said, especially if you need to straddle the old business and the
Greg:new business at the same time.
Greg:Sometimes, in my opinion, uh, call it and start again.
Greg:send back the capital if there's any left and, reset the cap table.
Greg:But again, that's another story.
Greg:We'll get there, in another episode.
Greg:Julia, thanks for joining us.
Greg:Congrats on your pivots worked clearly to a point where you had something
Greg:that someone else really valued, which was Handshake, and, uh, I hope
Greg:you enjoy your time working there, and that business gets an exit, so
Greg:you get one too, 'cause there's gotta be something besides the heartache
Greg:and tears of being an entrepreneur
Julia:I know we're done, but I will say uh, in just, one other piece of advice
Julia:is to build something you love.
Julia:build something that you are so passionate about that no matter what the
Julia:exit is or, what happens, you enjoyed it, and you are excited about it,
Julia:It wasn't a slog.
Julia:Like, it was something that, you were lucky enough to be able to show up
Julia:to work every day and do something meaningful that filled your bucket.
Julia:And so I think, yeah, the other advice that I would give for an entrepreneur
Julia:is set yourself up to be building something that you can endure.
Greg:Yeah, that's great advice, Julia.
Greg:'Cause that's what gets you through those troughs
Julia:it i- it is so hard.
Julia:the odds are against you, and you are…
Julia:It is the number of obstacles, the number of nos that we had from
Julia:customers who said that they were not interested in what we were building.
Julia:but do you wake up every day and you're like, "Oh my gosh, I get to build
Julia:something that I'm passionate about, that I have a vision around, with a set of
Julia:people that I really admire and think are special, and try to make a difference"?
Julia:Like, I think that's, energizing and fulfilling and what lets you keep going.
Julia:and so I think if you can…
Julia:like, what a gift to be able to be in a position where you get
Julia:to create that for yourself.
Julia:So that's the other thing.
Julia:Like, it's…
Julia:if, as an entrepreneur, yeah, if you can carve out or, you know, decide that
Julia:you wanna build something that you will enjoy building and that is meaningful
Julia:to you, it's just a massive competitive advantage and really helps prevent some of
Julia:the burnout that, that I think is, that,
Greg:That is inevitable or potentially inevitable anyway.
Greg:Yeah, yeah.
Greg:No, l- l- uh, let's end there.
Greg:That is great advice.
Greg:thanks, Julia, for joining us on "Super Companies."
Julia:Thanks so much.
Julia:Bye
