Built 2 Scale

What does it take to build an AI-native construction company?

This week, Matt takes the host’s chair to interview Scott Wilcox about the thinking, systems, and decisions behind W.—and his ambition to build in half the time at half the cost.

Scott explains why becoming more efficient at managing construction only gets you so far. If you improve by 30% but control just 10% of the cost, the overall improvement is 3%. That thinking is driving a deeper look at vertical integration, manufacturing, and the entire supply chain.

Fresh from visiting factories in China, Scott shares what he learned about where building products come from, how costs accumulate between manufacturers and job sites, and the opportunity to get closer to the source.

Back inside the business, they explore how Scott uses AI agents, turns repeatable work into reusable skills, builds construction programmes through conversation, and decides which software to buy and which to build.

They also discuss the human side: protecting time for judgment, relationships and craftsmanship, while keeping site teams focused on quality and safety.

In this episode:

• What makes a construction business AI-native.
• The connection between AI and vertical integration.
• Lessons from Chinese factories and construction supply chains.
• Practical starting points for builders adopting AI.
• Turning business processes into reusable skills.
• Bringing technical capability in-house.
• Building a company that can scale beyond its founder.
• Why quality lives at the interfaces.
• Wearables, AI agents, and a future where technology works quietly in the background.

Built to Scale is Scott Wilcox (W.) and Matt Perrott (BuildPass), building in public at the intersection of construction and AI.

What happens when intelligence becomes abundant, the cost of cognitive work trends towards zero, and humans themselves become the unpredictable variable inside increasingly autonomous systems?

In Episode 6, Scott and Matt explore what this means for the future of work, company design and the products businesses should be building.

The argument is not that humans have no value. It is that in some systems, unstructured human intervention can increasingly add cost, variance and risk. If machines can perform more of the work reliably, the question changes from “How do I make this person more productive?” to “How do I build the system that produces the outcome?”

That leads to the idea of the outcome machine: instead of repeatedly hiring people to perform the same task, combine deep domain expertise with AI-native system builders and encode that knowledge into a machine that gets better every time it runs.

They discuss why every company needs to put a harness around frontier technology and point it at its end product, why outputs are being commoditised while outcomes remain valuable, and why the real competitive advantage may sit between an increasingly powerful intelligence layer and the customer.

For construction, that leads somewhere much bigger: architecture and engineering becoming dramatically cheaper to produce, fragmented disciplines beginning to converge, computational design replacing static drawings, and construction evolving towards software + manufacturing.

Also covered: software factories, autonomous agents, AI spending disasters, Tesla and Cybercabs, self-maintaining physical products, the growing scarcity of skilled trades — and OnlyFans for electricians.

The question underneath the whole episode:

If the cost of intelligence keeps falling, what should your company actually become?

Built to Scale is Scott Wilcox (W.) and Matt Perrott (BuildPass) building in public at the intersection of construction and AI.

Follow Scott → / scott-wilcox-41bb2538

Follow Matt → / perrott

Structured AI is building an AI-native quality layer for construction: agents that can read complex drawing sets and 3D models, find design and coordination issues, and increasingly act inside Revit to help fix them. Their thesis is simple: QA/QC is still too manual, too inconsistent and too dependent on people catching everything under time pressure.

In this episode, Raymond Zhao and Ben Remy unpack how Structured AI is approaching that problem technically — combining software engineers with experienced architects and engineers, tightly managing agent context, and embedding domain knowledge so the system understands not just that something is wrong, but why it is wrong and how it should be resolved.

The conversation then moves beyond QA. Structured AI is already giving its Revit agent permission to make isolated changes to models, and sees a path from identifying errors, to fixing larger sub-problems, to eventually generating more of the building itself. In that future, the QA layer becomes even more important: if agents are designing and building, another agent still needs to verify the output.

A discussion about the evolution from fragmented construction documentation toward agent-generated, machine-checked buildings — and what it takes to make AI genuinely useful in the physical world.

Chapters:

0:00 Welcome — Ben & Raymond from Structured.ai

0:26 The mission: an AI quality layer for construction

2:25 Raymond’s path — architecture to Oxford stats

3:14 The fighter-jet origin story

6:05 Inside Y Combinator — solving the right problem

9:05 What is a forward-deployed engineer?

10:45 One of the great problems in construction

12:37 Under the hood: the giant vector soup

14:28 Context management is everything

17:47 The write side — fixing designs from first principles

20:42 Outcome as a Service

22:05 Who goes first on AI-generated design?

23:17 Productizing a messy industry

24:08 Biggest bottlenecks: scaling and talent

27:51 Should software set the standard?

29:36 Winning over the 40-year veterans

31:58 Agents checking agents — hallucination management

34:47 Closing thoughts + who should reach out

Structured.ai → https://getstructured.ai

Built to Scale is Scott Wilcox (W.) and Matt Perrott (BuildPass) building in public at the intersection of construction and AI.

Follow Scott → https://www.linkedin.com/in/scott-wilcox-wilcoxbuilt/

Follow Matt → https://www.linkedin.com/in/mattperrott/

The first ever live Built 2 Scale, streaming at 6am for the Australian drive to work and 3pm for the Austin drive home. Scotty and Matt dig into open source models like Kimi closing the gap on the frontier labs and what that does to cost per token for every business downstream, then get into the coming HR nightmare of token spend as a staff member four months into a European summer ranks number two in company token usage. They cover the shift from token maxing to LLM routing, why work is now completely mobile and completely voice, Granola landing on the Apple Watch and BuildPass following with Superite, and the 45 minute hackathon that produced Tinder for Talent, a LinkedIn scraping recruitment app so creepy Scotty's wife told him to delete it. Plus: live audience questions on AI glasses for site defect walks, whether founders will spend more time talking than typing, and how James Hird could use AI to fix Essendon.

Built 2 Scale | Season 2 Episode 5

TIMESTAMPS:

0:00 First Ever Live Built 2 Scale: Confirm or Deny the Bed in the Office

2:17 LeBron to the 76ers vs Jacinta Allan Leaving Victoria

4:00 Open Source Models Catch Up: Kimi, Fable Going Offline and Back

4:50 Good for Consumers and Businesses, Bad for the Frontier Labs

6:20 Token Maxing to Agent Architecture: Mapping Models to Your Org Chart

8:13 LLM Routing and the Uber Eats Milkrun Playbook for AI Pricing

11:46 The Token HR Problem: Staff Member in Italy Ranks Number Two on Spend

13:00 Do You Get Tokens on Annual Leave? Personal Use, Side Hustles and JD Perks

15:29 Tool of the Week: Scotty's Three Hour Voice Session Building the WOS

18:03 How to Frame It: Ask the AI to Interview You Until It Finds the Real Problem

19:35 Matt's 20 Minute Drive Home Debrief Built as a Claude Project

22:30 Whisper Flow and Why Typing Is the Dumbest Bandwidth You Own

23:30 Granola Hits the Apple Watch: The Passive Note Taker Moment

25:26 Breaking News: Superite Coming to the Apple Watch (Hit the Gong)

28:12 Site Diaries, Defect Walks and Pre Daily Briefings From Your Wrist

31:47 Building in Public: The 45 Minute Hackathon With an AI Engineer Per Team

36:03 Tinder for Talent: Swipe Right, Auto Message, Wife Says Delete It

39:11 Why Software Now Needs Tight Feedback Loops and Deep Verticals

40:34 Modular Software, MCP Servers and Why You Shouldn't Build Your Own CRM

43:44 Two Businesses Converging: The Builder Gets Technical, the Software Team Goes On Site

46:00 Outcomes as a Service: Coming for the Labour Bill, Not the Software Bill

47:31 Audience Q: How Far Away Are AI Glasses Capturing Defects on Site?

49:27 Scotty's Take: Instructions in the Glasses Beats Live Defect Detection

52:00 The World Model Layer: Live Site Data Rewriting the Schedule in the Background

53:12 Construction's Memory Problem: Everyone Goes Home and the IP Leaks Out

55:48 Audience Q: How Would James Hird Use AI to Fix Essendon?

56:54 Audience Q: Will Founders Talk More Than They Type? (Lovable Becomes Lava Bulb)

57:59 The VTT Hack: Flag Your Voice to Text So No One Judges Your Grammar

59:10 Tongue Controlled Mouse: The Weirdest Input Device Yet

This Episode Covers:

• Open source models like Kimi closing on the frontier and what that means for cost per token across every business
• Why hardware getting better and models getting smaller were always going to cross over in the middle
• Moving from token maxing to real agent architecture: mapping models to roles in your org chart by task type
• LLM routing and why not every request needs the most expensive model on earth
• The coming HR fight over tokens: annual leave, personal use, side hustles and tokens as an employment perk
• Voice as the new interface: Scotty's three hour session across a run, a gym, a sauna and a car producing a real artifact
• Framing voice sessions properly by asking the AI to interview you until it finds the actual bottleneck
• Matt's Claude project that turns the 20 minute drive home into a daily debrief and to do list
• Granola launching on Apple Watch and BuildPass bringing Superite to the wrist for site managers
• The 45 minute hackathon format: every team gets an AI engineer, lists ten pain points, ships a working prototype
• Tinder for Talent: the LinkedIn scraping recruitment prototype that replaced a $20K recruiter fee
• Why software companies need modular products, MCP access and deep customer feedback loops instead of point solutions
• AI glasses on construction sites: capture and processing are solved, the real time guidance layer is roughly six months out
• Construction's memory and communication problem as the biggest leak in the industry

KEY INSIGHTS:

• Open source is a consumer win and a frontier lab problem: Cheaper models plus better consumer hardware means prices have to come down. The only losers are the labs subsidising inference
• Token spend becomes a line item next to headcount: Finance teams already track salaries. Companies like RAMP are tracking AI spend. Expect role cost to mean salary plus agent cost
• Route the request, don't max the model: Sending a basic task to the most expensive model is the AI equivalent of Uber Eats VC subsidies. That era ends the moment margins matter
• Work is now mobile and voice first: A three hour session across a run, a gym and a car produced a foundational business artifact. The thumb to keyboard bandwidth is the bottleneck, not the thinking
• The Apple Watch finally has a real use case: Passive capture is less awkward than a phone on the table and frees your hands on site. Voice input is the killer feature, not health tracking
• Hackathons are 80/20 machines: Forty five minutes gets you a working prototype and a real idea. The last 20% takes 80% of the time, but you can't invest in an idea nobody has proven yet
• Don't rebuild the wheel, build what makes you different: A construction business shouldn't build its own CRM or payroll. Build the workflows unique to you on top of infrastructure that already exists
• Outcomes as a service is the next software frontier: Safety paperwork is hard to sell, finance is easier, winning tenders sells itself. The pitch is converting labour bills into tokens
• Instructions in the glasses beat defects in the glasses: Training vision models to spot a 5mm tolerance issue is expensive. Scrolling the SOP into a tradesman's eyeline scales quality across a workforce today
• The site as a world model: Live capture doesn't just flag defects, it feeds a system that reruns the schedule, spots downstream clashes and reschedules deliveries in the background
• Construction's leaky bucket: Everything seen and heard on site walks home at 5pm and never reaches tomorrow's crew. Any observation not shared to the project brain is lost IP
• Curious beats credentialed: The people getting the most out of these tools aren't waiting to be told what AI can do, they're assuming it can and finding the edges the hard way

About Built 2 Scale:

Weekly analysis of artificial intelligence business strategy, robotics adoption, and technology market dynamics. Hosted by Scott Wilcox (Australia) and Matt Perrott (Austin), providing actionable insights for founders, investors, and technology executives navigating the AI transformation.

Subscribe for AI business intelligence: /@built2scale

Available on Spotify | Apple Podcasts

#AI #OpenSource #Kimi #Claude #Fable #TokenSpend #LLMRouting #VoiceAI #Granola #AppleWatch #Hackathon #BuildPass #Superite #AIGlasses #Construction #Built2Scale #LiveStream #TechNews

How humans + tokens spark a lost-crafts renaissance

For two hundred years the master builder got fragmented — one person who could read a site, command a trade and run a job split into site manager, contracts administrator, project coordinator, estimator, fifteen roles deep.

That fragmentation wasn’t craft. It was admin: the tax you pay to run a build in the twenty-first century.

This week Scott and Matt make the case that tokens are about to pay that tax for us — and what’s left when the admin burden lifts is the craft itself. They go deep on the operating system underneath it: BuildPass’s move to four product lines — Win, Pay, Build, Connect — absorbing the fragmentation so one builder can sit back in the cockpit and give the verdict that matters: approve, challenge, or deny.

Then the human side — “human plus token,” hiring for agency over résumés, and why keeping real people in the building still matters. Plus the Traidy launch and raising the floor for 1.3 million trades, finishing your site admin by voice on the drive home, and a live demo of the open-source glasses that put a master’s guidance in your ear while you lay the brick. The robots aren’t coming to replace the bricklayer. They’re coming to make the bricklayer a master again.

Scott and Matt are back after a month off — and a new daughter, a new W. hire from the podcast itself, and a 22-minute plan to solve Australia. Matt coins "knowledge debt," Scott coins "tradesmen are skills files with dexterity," and they propose a hackathon that pairs a bricklayer with a developer. Plus Codex Goals running for four days unattended, AI cameras landing in Austin, and Scott drops the full Scotty Algorithm (DMICSA) as a public skill file. The world model, on a site.

Scott Wilcox and Matt Perrott break down the two weeks that took W.Build's invoice processing from $5–$10 per invoice to half a cent — a thousand times reduction — and what it signals for every construction business trying to figure out where agents fit.

In this episode:

Chapters / Timestamps

Mentioned / Referenced

Hosts

Built2Scale is back — and everything's different.

No more AI news roundups. From now on, every episode documents the real, messy, building-in-public journey of W. (luxury home builder, $30M+ revenue, ~20 people) and BuildPass (construction tech startup, 1,000+ customers across 4 countries) as they converge into one platform.

This week: Scott onboarded his entire company onto Claude in a single live session — $30/person/month, every tool connected, nobody left behind. Within days, his team stopped asking him questions and started asking the AI instead. ("Imagine if I spoke that eloquently at work.")

Matt discovered Paperclip — a tool that lets you run an agent-first company from your terminal. He hired a CEO agent, approved new hires through a board approval workflow, and watched autonomous agents start building a business. Then he left it running overnight and woke up to a rate limit error. The bill is TBD.

Also in this episode:

→ How to map agents against your org chart (Scott built a 20-page architecture in one weekend)

→ Brian Armstrong's "reverse prompting" approach at Coinbase

→ Scott's attempt to build Palantir for construction in 4 hours on Replit

→ Jensen Huang's two-question test for deciding what's worth working on

→ The WVC announcement — a Y Combinator-style accelerator for trades

Tools mentioned: Claude, Paperclip, OpenClaw, Replit, Monday.com, Granola, BuildPass, Notebook LM

Hosts: Scott Wilcox (@scottjwilcox) and Matt Perrott (@mattperrott)

In this episode, Scotty and Matt dive deep into Elon's limiting factor philosophy borrowed from physics (just remove the bottleneck, move to the next one, repeat), explore how they scored their own businesses across demand, authority, distribution, margin, talent, systems, capital, and CEO attention constraints, and reveal why Matt spent 80% of his time on US distribution while his LinkedIn posts generating hundreds of thousands of impressions fell off entirely. They dissect the paradox of intelligent people refusing AI tools due to ego protection, analyze Scotty's FebFast business sprint that turned February into four focused weekly deep dives, and demonstrate how Matt's AI bot now delivers five daily wins and five risks from every customer communication automatically. Plus: Why Aaron deleted one on ones to focus on building cutting edge products, the consumacute framework applied to business planning (learning holidays vs core revenue focus), and how over communicating in public channels builds the best AI training data for organizational knowledge.

Built 2 Scale | Episode 36

TIMESTAMPS:

0:00 Elon's Limiting Factor Philosophy: Just Remove The Bottleneck

2:55 Matt's Velocity Problem: Unblocking Everything But Not Focusing Right

4:47 Weekly Top 3 Wins and Losses Email: 1500 Reads Like Jensen

8:27 Moving All Communication to Public Channels For AI Knowledge Base

11:00 The Paradox: Intelligent People Refusing AI Due to Ego Protection

14:16 Aaron 10x Faster With AI Than Average Joe Without It

15:04 Staff Member Goes From Anti AI to Best Work in 48 Hours

15:13 The Eight Limiting Factors: Demand, Authority, Distribution, Margin, Talent, Systems, Capital, CEO Attention

17:05 Matt Scores BuildPass: Distribution is Number One Bottleneck

18:53 The LinkedIn Post Problem: 80% on US Sales, 0% on Organic Inbound

20:28 Scotty's FebFast Sprint: Four Week Business Bootcamp Focused on One Thing Weekly

26:23 Founder Deep Dives: Rotating Through Functions to Unpack Assumptions

27:00 Scotty's Demand Solution: Three Hour Blocks Building Whale Outreach Product

31:47 Aaron Deletes One on Ones to Build Cutting Edge AI Native OS

33:38 Delete The Requirement: Inspired Teams Don't Need HR Meetings

35:24 Burnout Comes From Not Winning: Work 12 Hour Days if You're Succeeding Weekly

36:21 The Bot Shamed Me: You're Going on a Learning Holiday Not Working

40:05 Designing Systems For Repeatability: Teaching Tez Context and Memory

42:51 Pumping Context Into Notion So Organizational Knowledge Doesn't Escape

45:38 Thought Spam Leadership: 50 After Hours Artifacts Compacting Four Hour Deep Thinks

47:02 Signal vs Noise: AI Compressing Business Data Into Actionable Takeaways

48:28 Matt's Daily Five Wins Five Risks Email From All Customer Communications

This Episode Covers:

• Elon's limiting factor philosophy from physics applied to business: Focus only on biggest bottleneck, remove it, move to next
• The eight business limiting factors: Demand, authority, distribution, margin, talent, systems, capital, CEO attention
• Matt scoring BuildPass as distribution constrained while spending 80% time on US sales but 0% on LinkedIn organic inbound
• Why intelligent people refuse AI tools due to ego protection around identity tied to craft mastery
• Aaron going 10x faster with AI than average Joe proving skill amplification not replacement
• Scotty's FebFast business sprint: Four weeks with whole company focused on new business, design, procurement, onsite velocity
• Why Aaron deleted one on ones to focus on building cutting edge AI native construction OS instead of playing HR babysitter
• Burnout comes from not winning not from working hard: Grand final winners aren't tired, losers lying on ground are
• The bot shaming Scotty's learning holidays: Reframing international travel as conscious consumption not core business work
• Designing repeatability with Tez: Building context memory through Obsidian notes synced to Mac Mini agent access
• Matt pumping Claude Code context into Notion documents so organizational knowledge doesn't escape conversation threads
• Thought spam artifacts: Scotty sending 50 after hours leadership emails annually compacting four hour deep thinking sessions
• Over communicate everything in public channels: AI makes signal from noise so never withhold information assuming irrelevance
• Matt's automated daily email: Five wins and five risks extracted from all customer communications across email, calls, chat, recordings

KEY INSIGHTS:

• Elon's limiting factor lens: Stop firefighting everything, identify single biggest bottleneck, focus entire attention removing it, move to next. Velocity without focus creates CEO as bottleneck
• The eight constraint scorecard: Demand, authority, distribution, margin, talent, systems, capital, CEO attention. Score yourself honestly then time audit whether calendar matches priority
• LinkedIn opportunity cost: Matt 80% on US sales enablement, 0% on posts generating hundreds of thousands impressions. Distribution bottleneck requires multi channel time allocation audit
• Intelligence paradox: Smart people tied to craft identity resist AI tools. Dumber curious people just trying stuff having more impact than experts assuming AI limitations
• Skill amplification not replacement: Aaron 10x faster with AI than average Joe proves expertise compounds with tools. Craft knowledge guides AI direction, taste still matters
• FebFast sprint compression: Whole company four week bootcamp with weekly themes (new business, design, procurement, onsite) creates January momentum and shared mission alignment
• Delete HR requirements: Aaron ditching one on ones to build cutting edge products. Inspired teams drawn to mission not needing founder babysitting interpersonal conflicts
• Winning prevents burnout: Grand final winners running around with cup aren't tired, losers on ground are. Hard work with weekly wins sustains 12 hour days for years
• Learning holiday reframe: Bot shaming Scotty for international travel disguised as business development. Conscious consumption acknowledges exploration not convincing self it's core work
• Repeatability design investment: Teaching Tez context through Obsidian notes, swimming lesson preferences, son's age. Unsexy upfront work enables Neo in Matrix moment later
• Notion as escape valve: Claude Code pumping insights into shared documents prevents context compacting memory loss. Artifacts must exit personal AI into organizational system
• Thought spam leadership value: 50 after hours artifacts annually from CEO deep thinking compacts knowledge other executives never get. Four hour sessions into one pagers scales insights
• Over communication wins with AI: Signal vs noise problem solved by AI compression. Never withhold information assuming irrelevance, let others manage notifications and search later
• Automated daily intelligence: Five wins five risks from all customer communications (email, calls, chat, recordings) surfaces complaints and victories leaders never hear otherwise

About Built 2 Scale:

Weekly analysis of artificial intelligence business strategy, robotics adoption, and technology market dynamics. Hosted by Scott Wilcox (Australia) and Matt Perrott (Austin), providing actionable insights for founders, investors, and technology executives navigating the AI transformation.

Subscribe for AI business intelligence: /@built2scale

Available on Spotify | Apple Podcasts

#AI #Elon #LimitingFactor #Bottlenecks #BusinessStrategy #FebFast #Claude #OpenClaw #Tez #Distribution #Focus #Burnout #AIAgents #OrganizationalKnowledge #Built2Scale #FounderMode #TechNews

BUILT 2 SCALE | TOOL OF THE WEEK | Episode 35

Every week, Matty and Scotty break down the strategies, frameworks, and tools that separate the world's best builders from everyone else.

This week: Consumecute, the workflow that turns mindless consumption into active execution.

The Trigger: Am I Consuming or Building?

Scotty's 2026 resolution: check if you're consciously consuming or mindlessly scrolling. Conscious consumption builds. Mindless scrolling is gluttony—input without output.

Claude Cowork Challenge

Matty's challenge: think about your most time consuming process. Try automating it with Claude Cowork. You'll be surprised what's possible.

Consumecute: Capture Your Consumption

Export your Apple Podcast history, drop into ChatGPT for takeaways, plug into Monday.com or Notion. Build training data for AI to run your decisions through everything you've consumed.

Connect Productive and Consumption Tools

Plug in emails, Teams messages, everything. Ask AI: "Based on how I'm communicating, is this aligned with my values?" Stop living separate lives between consumption and productivity.

The Audible Experiment

Scotty condensed 110 Audible books into one sentence: "A system is an incentive structure pointed at an outcome." 44 books over 10 years into one formula. Now used at work: vision (CEO), incentives (humans), system (automation).

Zoom In and Out

Use your condensed sentence as North Star or zoom into details on specific topics. Capture data that matters. Fine tune your decision logic. This took 15 minutes with a few plugins and one automation.

The Takeaway:

Recapture your attention. Turn consumption into execution. Connect tools. Guilt free scrolling exists when pointed at an outcome.

Are you consciously consuming or mindlessly scrolling? What habit are you turning into leverage in 2026?

Keen to stay ahead? Subscribe to Built 2 Scale on YouTube (link in comments)