For years, software companies followed the same blueprint:

Raise venture capital. Hire aggressively. Sell SaaS subscriptions. Scale headcount alongside revenue.

But AI is rewriting those rules.

In this episode of Verticals, Luke and Nic sit down with Zain Jaffer, who previously built Vungle into a business that was acquired for $780M before later reaching a $4B public valuation. After investing in startups as a VC, he's now back in the founder seat building Blazel, and unlearning many of the lessons that made him successful the first time.

One of the biggest takeaways from this episode is that AI-native agencies may look like the easiest businesses to build today, but they can quickly become some of the hardest businesses to scale.

Zain explains why replacing services with AI creates an attractive wedge, but can also trap founders in low-margin businesses that become difficult to escape.

They unpack some of the biggest debates in AI right now:

-Why the traditional SaaS playbook is starting to break

-The hidden risks of AI-native agencies

-When founders should build on OpenAI, and when they should build their own models

-Why distribution is becoming a stronger moat than product alone

-How AI is changing hiring, product development, and company building

🎟️ TICKETS DROPPED. Vertical Software Summit — 5-billion-dollar Vertical AI CEOs confirmed to speak. Snag a ticket before they're gone.

👉 FOLLOW ZAIN JAFFER

WEBSITE ➡️ https://zainjaffer.com/

LINKEDIN ➡️ https://www.linkedin.com/in/zainjaffer/

👉 FOLLOW LUKE SOPHINOS

X ➡️ https://x.com/lukesophinos

LINKEDIN ➡️ https://www.linkedin.com/in/lukesophinos/

NEWSLETTER ➡️ https://www.newsletter.lukesophinos.com/

👉 FOLLOW NIC POULOS

X ➡️ https://x.com/picnoulos

LINKEDIN ➡️ https://www.linkedin.com/in/npoulos/

NEWSLETTER ➡️ https://insights.euclid.vc/

🎬 TIMESTAMPS:

00:00 Intro

01:30 Building an AI marketing platform from scratch

08:05 The future of AI-native marketing teams

17:14 Why data flywheels are the new moat

20:31 Choosing the right ICP in the AI era

25:36 The hidden trap of AI-native agencies

33:25 Human labeling, evals, and building better AI

37:42 Should every startup build its own model?

42:10 Are OpenAI and Anthropic worth a trillion dollars?

49:23 What an $800M founder had to unlearn about startups

When ChatGPT launched, investors had a simple answer for most AI startups:

"It's just a GPT wrapper."

But what if the biggest winners weren't the model companies?

What if the real opportunity was building products that deeply understood a specific customer, workflow, and industry?

In this episode of Verticals, Luke and Nic sit down with Adeel Khan, Founder and CEO of MagicSchool, one of the fastest-growing AI companies in education.

Adeel wasn't a software founder. He was a teacher, assistant principal, and school leader who saw the potential of generative AI firsthand and decided to build for educators. Just months later, MagicSchool exploded from zero to 1 million users in five months, entirely through organic growth.

Today, the platform is used by educators in nearly every school building in America and partners with districts representing one in five children nationwide.

They discuss:

-Why domain expertise became MagicSchool's biggest advantage

-How a non-technical founder built one of the fastest-growing AI companies

-The controversial decision to give the product away for free

-How they converted massive product usage into enterprise revenue

-Why GPT wrappers were more defensible than many investors believed

And Adeel shares his perspective on the future of education, the opportunities AI is creating for students and teachers, and why there has never been a better time to build software for underserved industries.

If you're building in vertical AI, education, or trying to understand what separates lasting companies from short-lived AI hype, this episode is packed with lessons from one of the category's breakout founders.

🎟️ TICKETS DROPPED. Vertical Software Summit — 5-billion-dollar Vertical AI CEOs confirmed to speak. Snag a ticket before they're gone.

👉 FOLLOW ADEEL KHAN

LINKEDIN ➡️ https://www.linkedin.com/in/adeelkh/

👉 FOLLOW LUKE SOPHINOS

X ➡️ https://x.com/lukesophinos

LINKEDIN ➡️ https://www.linkedin.com/in/lukesophinos/

NEWSLETTER ➡️ https://www.newsletter.lukesophinos.com/

👉 FOLLOW NIC POULOS

X ➡️ https://x.com/picnoulos

LINKEDIN ➡️ https://www.linkedin.com/in/npoulos/

NEWSLETTER ➡️ https://insights.euclid.vc/🎬

TIMESTAMPS:

00:00 Intro

01:30 From school principal to AI founder

05:17 Launching MagicSchool during the ChatGPT moment

08:59 Why domain expertise matters more than ever

14:11 The fundraising challenges of a non-technical founder

20:13 How MagicSchool's product-led growth exploded

23:41 The free-to-paid strategy that unlocked enterprise sales

29:02 Transitioning from order-taking to real sales

34:07 The biggest lesson from scaling to millions of users

36:14 Why GPT wrappers became more valuable, not less

43:01 Competing with legacy education software companies

46:46 Is there still room to build in education AI?

48:05 Why education remains one of the biggest opportunities in software

Most founders think investors are looking for the perfect idea.

They're not.

At the earliest stages, investors are really trying to answer a much harder question:

Is this founder capable of building something extraordinary?

In this episode of Verticals, Luke and Nic sit down with Dan Teran and James Gettinger of Gutter Capital to unpack exactly how they evaluate founders before there's meaningful revenue, traction, or even a finished product.

They break down:

-Why resourcefulness beats pedigree

-What "founder-market fit" actually means

-How customer obsession separates great founders from everyone else

-Why evidence beats storytelling every time

-The founder traits they consistently see in breakout companies

🎟️ TICKETS DROPPED. Vertical Software Summit — 5-billion-dollar Vertical AI CEOs confirmed to speak. Snag a ticket before they're gone.

👉 FOLLOW DAN TERANLINKEDIN ➡️ https://www.linkedin.com/in/danteran

👉 FOLLOW JAMES GETTINGERLINKEDIN ➡️ https://www.linkedin.com/in/james-gettinger-452a228X ➡️ https://x.com/chess_is_ok

👉 FOLLOW LUKE SOPHINOSX ➡️ https://x.com/lukesophinosLINKEDIN ➡️ https://www.linkedin.com/in/lukesophinos/NEWSLETTER ➡️ https://www.newsletter.lukesophinos.com/

👉 FOLLOW NIC POULOSX ➡️ https://x.com/picnoulosLINKEDIN ➡️ https://www.linkedin.com/in/npoulos/NEWSLETTER ➡️ https://insights.euclid.vc/

🎬 TIMESTAMPS:

00:00 Intro

03:59 Why Gutter Capital was built differently

06:04 Resourcefulness beats pedigree every time

10:10 The traits that predict exceptional founders

17:12 How VCs really reference-check founders

28:17 The most surprising startups in their accelerator

34:52 The final advice every founder should hear

If you enjoyed this episode, drop a comment and hit subscribe so you don’t miss what’s next!

For years, venture capital forced founders into boxes.

You were either a consumer company or an enterprise company.

But some of the biggest companies in the world never followed those rules.

Today, the line between consumer and enterprise is becoming harder to see… and AI may be accelerating that shift.

OpenAI, Anthropic, and Google are moving faster than anyone expected. And for many startups, the biggest risk isn't competition from another startup, it's building directly into the AI kill zone.

In this episode of Verticals, Luke and Nic sit down with Nikhil Basu Trivedi, Co-Founder and GP of Footwork VC, to discuss how founders can build defensible companies in an era dominated by foundation models, and why the next generation of category-defining companies may need to win both consumers and enterprises at the same time.

They unpack some of the biggest questions facing founders right now:

-Is the line between consumer and enterprise disappearing?

-What actually creates a moat in the AI era?

-How do you avoid building directly into the kill zone of OpenAI and Anthropic?

-Which overlooked industries are ripe for AI transformation?

👉 FOLLOW NIKHIL BASU TRIVEDI

LINKEDIN ➡️ https://www.linkedin.com/in/nikhilbt

X ➡️ https://x.com/nbt

👉 FOLLOW LUKE SOPHINOS

X ➡️ https://x.com/lukesophinos

LINKEDIN ➡️ https://www.linkedin.com/in/lukesophinos/

NEWSLETTER ➡️ https://www.newsletter.lukesophinos.com/

👉 FOLLOW NIC POULOS

X ➡️ https://x.com/picnoulos

LINKEDIN ➡️ https://www.linkedin.com/in/npoulos/

NEWSLETTER ➡️ https://insights.euclid.vc/

🎬 TIMESTAMPS:

00:00 Intro

03:16 The venture thesis behind Footwork VC

05:11 Why the best companies win consumers and enterprises

13:05 Should founders focus or expand faster in AI?

24:40 Network effects, moats, and what still matters in AI

31:14 The biggest opportunities in 2026

35:12 Avoiding the AI kill zone

40:27 What makes a vertical AI market worth pursuing?

42:38 The hottest AI companies in 202

49:17 Final lessons for founders building in the AI era

The biggest problem in AI isn't building the product.

It's staying relevant six months later.

Models improve every month. New competitors appear every week. And switching costs are lower than most founders want to admit.

So what actually creates a moat?

In this episode of Verticals, Luke and Nic sit down with Kyle Lui from Bling Capital to break down the new rules for product-market fit, market sizing, and defensibility in the AI era.

They also cover:

-Why AI-native companies are growing faster than SaaS ever did

-Why some of the best opportunities now look “too niche” for traditional VC

-How AI is changing org charts, hiring, and management itself

If you are building, investing, or trying to understand where vertical AI is actually headed, this is a sharp tactical breakdown of what matters now(and what probably won’t matter anymore).

👉 FOLLOW KYLE LUI

LINKEDIN ➡️ https://www.linkedin.com/in/kylelui

👉 FOLLOW LUKE SOPHINOS

X ➡️ https://x.com/lukesophinos

LINKEDIN ➡️ https://www.linkedin.com/in/lukesophinos/

NEWSLETTER ➡️ https://www.newsletter.lukesophinos.com/

👉 FOLLOW NIC POULOS

X ➡️ https://x.com/picnoulos

LINKEDIN ➡️ https://www.linkedin.com/in/npoulos/

NEWSLETTER ➡️ https://insights.euclid.vc/

🎬 TIMESTAMPS:

00:00 Intro

03:26 The frameworks top investors use to spot breakout startups

07:19 What product-market fit actually means now

13:12 Why “AI services” are changing software economics

18:05 The hardest question in AI

24:36 The hidden advantage of routing across multiple models

31:04 Are AI models becoming a commodity?

35:47 How AI companies tap into labor budgets

39:16 Why “too niche” markets might become massive in AI

40:23 Will vertical AI become winner-take-all like SaaS?

45:14 The biggest mistake investors are making right now

52:21 The niche AI market that surprised even their LPs

If you enjoyed this episode, drop a comment and hit subscribe so you don’t miss what’s next!

Everyone thinks AI will change software through chat.

But the biggest shift might actually happen through voice.

Because when humans stop typing and start talking to AI naturally, the amount of context, emotion, and real-world data flowing into software explodes.

And that changes everything.

In this episode of Verticals, Mike Droesch from Bessemer Venture Partners breaks down why voice AI is becoming one of the most important layers in vertical software.

And we also go deep on what is actually happening inside the Voice AI boom:

-Why voice agents are finally good enough to use in production

-How AI voice infrastructure became a massive opportunity

-Why regulated industries like healthcare and insurance are adopting faster than expected

They also unpack the biggest debates in AI right now:

-Is Voice AI becoming a commodity?

-Where will defensibility actually come from?

-Why evals, workflow ownership, and data flywheels matter more than ever

If you are building in AI, SaaS, healthcare, fintech, or customer support, this episode is one of the clearest breakdowns of where Voice AI is headed next.

🎬 TIMESTAMPS

00:00 Why Voice AI could transform software forever

03:15 The infrastructure powering the Voice AI boom

06:04 Why healthcare and finance are adopting Voice AI fast

11:15 The hidden moat behind Voice AI platforms

15:34 Why humans still matter in AI voice workflows

19:45 Voice AI use cases nobody saw coming

24:26 The real future of AI interfaces

29:11 Which Voice AI markets are overcrowded… and which aren’t

34:41 How AI pricing models are changing software economics

37:53 What investors actually look for in AI startups now

41:32 The biggest mistake Voice AI founders make

42:55 Why owning the entire workflow matters

📃 TOPICS

1. Voice AI & AI Voice Agents

2. Vertical AI & Vertical SaaS

3. AI Infrastructure & Developer Platforms

4. Healthcare AI, Fintech AI & Regulated Industries

5. The Future of AI Interfaces & Conversational Software

Everyone wants to build an AI company right now.But most founders are asking the wrong question.The real challenge isn’t building AI.It’s building something defensible once everyone else has access to the same models.In this episode of Verticals, the tables turn.Nic joins Vignesh Ravikumar and Brendon Schmidt from Sierra Ventures for a deep dive into what it actually takes to build and invest in vertical AI today.Nic shares lessons from nearly 15 years investing in vertical software and AI and explains why the biggest opportunities in AI aren’t just about models… they’re about workflows, distribution, and defensibility.They unpack the biggest shifts happening right now:-Why vertical AI is evolving faster than SaaS ever did-How AI-native startups are attacking incumbents-Why “speed” alone is not a moatAnd they tackle the biggest debate in tech right now:Is SaaS actually dying… or are we just watching a new generation of winners emerge?Watch until the end to find out!

#verticals #startup #ai #entrepreneurship #businessideas 👉 FOLLOW VIGNESH RAVIKUMARLINKEDIN ➡️https://www.linkedin.com/in/vravikumar1

👉 FOLLOW BRENDON SCHMIDTLINKEDIN ➡️https://www.linkedin.com/in/brendonschmidt/👉 FOLLOW LUKE SOPHINOSX ➡️ https://x.com/lukesophinosLINKEDIN ➡️ https://www.linkedin.com/in/lukesophinos/NEWSLETTER ➡️ https://www.newsletter.lukesophinos.com/👉 FOLLOW NIC POULOSX ➡️ https://x.com/picnoulosLINKEDIN ➡️ https://www.linkedin.com/in/npoulos/NEWSLETTER ➡️ https://insights.euclid.vc/🎬 TIMESTAMPS:00:00 Intro03:01 The shift from vertical SaaS to vertical AI08:18 How AI is changing the founder profile12:24 Why customer discovery matters more than ever18:55 How vertical AI startups beat incumbents28:14 Why “speed” is NOT a moat34:28 The rise of AI services businesses39:35 The vertical AI sectors investors want most42:39 How founders should position their company49:25 Is contrarian investing still alive in venture?58:59 The biggest mistakes founders make early on01:09:12 Is SaaS actually dying?If you enjoyed this episode, drop a comment and hit subscribe so you don’t miss what’s next!

In high-stakes industries, 99% accurate is the same as wrong.

That last 1% is the entire business.

In renewable energy, a single mistake can delay projects by months… or kill them entirely.

That’s exactly what Jacob Sandry saw while deploying over $3B into renewable energy projects at Goldman Sachs.

60% of projects were delayed. 30% were canceled altogether.

Not because the technology didn’t work… but because the process was chaos.

In this episode of Verticals, Jacob breaks down how Euclid Power is building the operating system for renewable energy — using AI, structured workflows, and human expertise to solve the “last messy mile” of energy development.

They unpack why vertical AI works differently in regulated, mission-critical industries:

-Why AI alone isn’t enough when mistakes have real consequences

-How humans-in-the-loop become a competitive advantage

-Why trust, compliance, and domain expertise matter more than model quality

They also go tactical on what is working right now:

-Why services + AI can outperform pure SaaS in vertical markets

-How to price against labor budgets instead of software budgets

-Why unstructured data is the biggest opportunity in enterprise AI

If you are building in AI, energy, healthcare, construction, or any industry where accuracy and trust matter more than speed, this episode is a masterclass in vertical AI.

👉 FOLLOW JACOB SANDRY

LINKEDIN ➡️ https://www.linkedin.com/in/jacobsandry/

👉 FOLLOW LUKE SOPHINOS

X ➡️ https://x.com/lukesophinos

LINKEDIN ➡️ https://www.linkedin.com/in/lukesophinos/

NEWSLETTER ➡️ https://www.newsletter.lukesophinos.com/

👉 FOLLOW NIC POULOS

X ➡️ https://x.com/picnoulos

LINKEDIN ➡️ https://www.linkedin.com/in/npoulos/

NEWSLETTER ➡️ https://insights.euclid.vc/

🎬 TIMESTAMPS:

00:00 Intro

01:55 The hidden bottleneck nobody talks about

05:01 From “alternative energy” to a $3B investment strategy

10:13 Why AI services work better than pure SaaS here

18:46 The multiplayer workflow behind billion-dollar projects

21:03 Why AI still struggles with the “last 1%”

31:37 Selling into industries where mistakes aren’t allowed

44:47 Is regulation actually a moat in AI?

47:20 Why energy could become the biggest AI opportunity of all

49:19 The long-term vision

If AI can do the work… why are we still valuing software like it only sells tools?And if the real opportunity isn’t software (but labor) are we about to underestimate the biggest companies of this cycle?In this episode of Verticals, Luke and Nic sit down with David Haber, General Partner at Andreessen Horowitz, to break down why vertical AI is fundamentally changing how businesses are built, and why this shift could be bigger than SaaS itself.David explains why AI-native companies aren’t just selling software — they’re replacing labor. And how that single shift unlocks markets that were previously too small to matter.They unpack the new playbook for vertical AI:-Why TAM is no longer software spend, it’s labor-How messy inputs like email, fax, and voice are becoming billion-dollar wedges-Why AI-native companies are turning services businesses into high-margin softwareThey also explore what actually creates defensibility in this new world:-Why owning workflows and outcomes matters more than ever-How proprietary data compounds into a real moat-Why market structure is more important than market sizeAnd they go deep on the biggest strategic question:If AI owns the workflow, the data, and the outcome… Do you even need to charge for the system of record?If you are building, investing, or trying to understand where AI is actually creating value, this is one of the clearest breakdowns of what comes next.👉 FOLLOW DAVID HABERX ➡️ https://x.com/dhaberLINKEDIN ➡️ https://www.linkedin.com/in/haber/👉 FOLLOW LUKE SOPHINOSX ➡️ https://x.com/lukesophinosLINKEDIN ➡️ https://www.linkedin.com/in/lukesophinos/NEWSLETTER ➡️ https://www.newsletter.lukesophinos.com/👉 FOLLOW NIC POULOSX ➡️ https://x.com/picnoulosLINKEDIN ➡️ https://www.linkedin.com/in/npoulos/NEWSLETTER ➡️ https://insights.euclid.vc/🎬 TIMESTAMPS:00:00 Intro01:28 Why the SaaS-AI shift is bigger than it looks07:09 Why “small markets” suddenly become massive10:06 The 3 ways to attack legacy software15:10 Why incumbents still have an advantage18:02 The real risk (and opportunity) in the AI era27:31 The messy inbox wedge (email, fax, voice)34:31 Why unstructured data is the biggest opportunity40:43 Voice, context, and the future of enterprise data44:15 The biggest mistakes vertical AI founders make45:28 The next wave: financial services, healthcare, industrial AI📃 Topics:1. Vertical AI vs SaaS2. AI replacing labor vs software tools3. System of record vs AI workflows4. AI services vs software margins5. Defensibility in AI startups6. TAM expansion through automation

Everyone says SaaS is dead.

Public software stocks are down. Multiples are collapsing. And AI is coming for everything.

But what if that’s not the full story?

In this episode of Verticals, Nic sits down with Morgan Livermore, Founder of Supercruise Capital, to break down what’s actually happening in the software market, and where AI fits into it.

Morgan explains why SaaS multiples have dropped below the S&P 500 for the first time in modern history, why the market may be overreacting, and why software isn’t dying... it’s evolving.

If you are building, investing, or trying to understand what AI actually means for software, this is a clear, grounded breakdown of where things are headed.

TIMESTAMPS:

00:00 Intro

01:03 Why founders are skipping venture capital

05:44 Is the “SaaS-pocalypse” real?

18:18 How AI is changing the startup landscape

21:48 Will AI replace SaaS or just extend it?

26:33 Can big companies adapt?

31:36 Why commoditization is coming for software pricing

37:39 What actually makes a company defensible now

44:15 AI agents, distribution, and the next growth channel

52:38 The real future of SaaS