What founders are actually building
before the analysts notice
Founder Pulse is a narrative trend forecast — built from hundreds of conversations with the people actually shipping code, raising rounds, and hiring engineers. We don't wait for the Crunchbase data to catch up. We listen to the people the data is about.
01 The five theses for 2026
Vertical AI eats horizontal SaaS
The 2010s playbook of "horizontal SaaS for everyone" is being unbundled by domain-specific AI agents that own a vertical end-to-end. Founders are no longer building tools — they're building digital workers.
Hardware is back, and it's American
For the first time since the late 90s, the most ambitious founders we talk to are building physical things — robots, semis, batteries, satellites. Capital follows ambition.
The energy x AI flywheel
AI compute needs power. Power needs investment. Investment needs returns from AI compute. Houston and Austin are stitching the cycle together in ways the coasts can't match.
Defense as the new growth sector
The most aggressive Series A founders we know aren't building consumer apps — they're building defense autonomy. The barrier to entry collapsed in 24 months.
Climate stops being charity
Climate-tech founders no longer pitch on impact. They pitch on margin. The decarbonization stack is now a real industrial market with real pricing power.
02 The conversations behind these theses
We don't summarize. We quote. Each thesis above is built from 30+ on-the-record interviews with founders, plus another 60+ background conversations with VCs, talent leaders, and ecosystem operators.
03 Geographic deep dives
Austin →
Eight neighborhoods, eight founder archetypes. Where the next Austin Origin Stories are being written.
ReadTexas →
Beyond Austin: the Houston energy-AI flywheel, DFW's defense renaissance, San Antonio's quiet cyber boom.
ReadUS Tech Centers →
Twelve cities, twelve cultures. Why the Bay Area still leads, why Boston still matters, why Miami matters more than people think.
ReadGlobal →
From Shenzhen's hardware density to Lagos's mobile-first AI experiments. Where the world's founders are placing their bets.
Read04 The evidence on the ground
A thesis is only as good as what's actually happening. Here's the public, verifiable backdrop each of the five rests on — the macro facts, not the anecdotes.
Vertical AI
The shift from selling software seats to selling completed work is the defining product change of the cycle — agents that own a workflow end-to-end rather than a dashboard a human still has to drive. The pricing question (per-seat vs. per-outcome) is where the real disruption shows up.
American hardware
The CHIPS Act and a broad reshoring push put real public capital behind domestic manufacturing. Central Texas is a direct beneficiary — Samsung's multibillion-dollar Taylor fab anchors a supplier ecosystem, and Tesla's Giga Texas pulled a robotics and battery talent pool to the region.
Energy × AI
Data-center electricity demand is the fastest-growing load on the Texas grid, and ERCOT's forecasts have been revised sharply upward. Texas pairs cheap land, an independent grid, and an energy industry that already knows how to finance big infrastructure — the structural advantage the coasts can't easily copy.
Defense
Venture funding into defense tech has climbed for several years running, led by autonomy and software-defined systems. The Pentagon's commercial-first units lowered the barrier for startups, and Austin already hosts a growing defense-and-dual-use cluster around Capital Factory and Army Futures Command.
Climate as industry
Inflation Reduction Act incentives turned decarbonization from a cost center into a financeable industrial market. Founders increasingly pitch unit economics — cost per ton, payback period — rather than impact, because the subsidies and offtake demand now make the math work.
05 The counter-case
A forecast that can't be wrong isn't a forecast. Here's the bear argument against each thesis — the way to read Founder Pulse honestly is to hold both.
Against Thesis 1
Agent reliability is still the bottleneck. Buyers who got burned by a hallucinating “digital worker” pull back hard, and outcome-based pricing only works when the outcome is verifiable. Plenty of “vertical AI” is a thin wrapper a foundation-model update can erase overnight.
Against Thesis 2
Hardware is brutally capital-intensive and slow. Fabs and factories take years and tolerate no software-style iteration. A funding-market tightening or a demand air-pocket strands a lot of ambitious atoms-not-bits companies before they reach scale.
Against Thesis 3
The flywheel runs on permits, transmission, and water — none of which move at startup speed. Interconnection queues are years deep, communities are pushing back on data-center siting, and an AI-demand slowdown would leave a lot of speculative power capacity looking for a buyer.
Against Thesis 4
Defense procurement is still a long, relationship-heavy slog, and budget cycles turn. A startup can win a flashy prototype contract and still die waiting for a program of record. The barrier to a first contract fell; the barrier to a durable business did not.
Against Thesis 5
Climate economics lean on policy that can reverse with an election, and many of these businesses are rate-sensitive and commodity-exposed. When subsidies wobble or capital gets expensive, “margin, not mission” gets tested fast.
06 Coming this quarter
- The East Austin AI Garage Tour — a walking-distance map of 38 AI infra startups in 1.2 sq mi
- Houston's Energy-AI Stack — five-part series on how Texas energy majors became hyperscaler customers
- The DFW Defense Cohort — interviews with 12 defense-AI founders and the primes hiring them
- Lagos to Nairobi — field reports from East Africa's mobile-first AI builders