Engineering Is Back

Engineering Is Back

Coding is figured out. Engineering is back. That's the shift, and it's already showing up in who excels and who gets stuck. Carrick Carpenter, Chief Technology and AI Officer at Aditi Consulting, sits down with Andrew Duncan for a peer conversation about what actually changes inside a services firm when AI rewires the delivery model. Aditi runs 2,000+ employees, 70+ Fortune 500 clients, three continents. Andrew runs Vertice Labs at the boutique end. Same tension, different scale. They go past "AI-native" positioning into the mechanics: why roles are collapsing, what a 90-pull-request engineer looks like in practice, why the people best at AI burn out fastest, and what Carrick got wrong in his first four months on the job. Topics * Why engineering is back and what that actually means * The 90-pull-request engineer and how to scale the pattern * Cycle time as the new delivery metric * The capacity paradox inside services firms * Full stack leaders and collapsing roles * What AI-native actually looks like at Fortune 500 scale * Burnout risk for AI-native teams * Outcome-based contracts and enterprise procurement * SaaS in the agentic era Guest Carrick Carpenter, Chief Technology & AI Officer at Aditi Consulting LinkedIn: https://www.linkedin.com/in/carrickcarpenter [https://www.linkedin.com/in/carrickcarpenter] Aditi Consulting: https://www.aditiconsulting.com [https://www.aditiconsulting.com] Resources mentioned Claude Code: https://claude.com/product/claude-code [https://claude.com/product/claude-code] OpenClaw: https://en.wikipedia.org/wiki/OpenClaw [https://en.wikipedia.org/wiki/OpenClaw] Gong: https://www.gong.io [https://www.gong.io] MCP: https://modelcontextprotocol.io [https://modelcontextprotocol.io] Salesforce: https://www.salesforce.com [https://www.salesforce.com] Slalom: https://www.slalom.com [https://www.slalom.com] Vertice Labs: https://www.verticelabs.io [https://www.verticelabs.io] Connect Andrew Duncan: https://www.linkedin.com/in/andrew-b-duncan/ [https://www.linkedin.com/in/andrew-b-duncan/] Vertice Labs: https://www.verticelabs.io [https://www.verticelabs.io] Subscribe Spotify: https://open.spotify.com/show/64yoHwDdhBd0vik2sev6xs [https://open.spotify.com/show/64yoHwDdhBd0vik2sev6xs] Apple Podcasts: https://podcasts.apple.com/gh/podcast/shipping-excellence/id1794200466 [https://podcasts.apple.com/gh/podcast/shipping-excellence/id1794200466] YouTube: https://www.youtube.com/@shippingexcellence [https://www.youtube.com/@shippingexcellence] RSS: https://media.rss.com/shippingexcellence/feed.xml [https://media.rss.com/shippingexcellence/feed.xml]

SaaS Isn't Dead. The Moat Just Moved.

SaaS isn't dead. UI is. The moat just moved. Assaf Weinberg (Co-Founder & CEO, SetSale) joins Andrew Duncan to dismantle the LinkedIn narratives about software in the age of AI: "code is free," "SaaS is dead," and the founder myth that hiring a developer is the bottleneck. Chapters * 00:00 The biggest limit on AI is what you believe it can do * 01:23 Building SetSale before AI tools existed * 04:31 Supastarter and the portal boilerplate origin * 06:47 Why the cost of software isn't zero * 07:43 The "I just need a developer" myth * 11:09 Every software company is a data transformation engine * 13:58 Why SaaS isn't dead * 17:08 Capture your data or AI can't help you * 22:20 The Austin distributor story: four days to ship * 27:34 Artists don't sketch in marble * 33:30 A developer's job is specification, not code * 44:13 The new bottleneck is customer adoption Topics * Why AI made code cheap but didn't make software cheap * The "I just need a developer" myth and why it was always a crutch * UI as commodity, data and process as the only durable moat * Why every software company reduces to a data transformation engine * The Austin distributor visit: notetaker, AI PM, Cursor, four days to ship * Six-week bets as the new unit of org change * What roles look like when designers don't make mockups and developers don't write code * The new bottleneck: customer adoption now lags engineering velocity Guest Assaf Weinberg, Co-Founder & CEO, SetSale https://www.setsale.ai/ [https://www.setsale.ai/] https://www.linkedin.com/in/assafweinberg/ [https://www.linkedin.com/in/assafweinberg/] https://x.com/assafweinberg [https://x.com/assafweinberg] Resources mentioned Sense and Respond by Jeff Gothelf and Josh Seiden: https://www.amazon.com/Sense-Respond-Successful-Organizations-Continuously/dp/1633691888 [https://www.amazon.com/Sense-Respond-Successful-Organizations-Continuously/dp/1633691888] Granola (AI notetaker): https://www.granola.ai/ [https://www.granola.ai/] Supastarter (Next.js SaaS boilerplate): https://supastarter.dev/ [https://supastarter.dev/] Cursor: https://www.cursor.com/ [https://www.cursor.com/] Connect Andrew Duncan: https://www.linkedin.com/in/andrew-b-duncan/ [https://www.linkedin.com/in/andrew-b-duncan/] Vertice Labs: https://www.verticelabs.io [https://www.verticelabs.io] Subscribe Spotify: https://open.spotify.com/show/64yoHwDdhBd0vik2sev6xs [https://open.spotify.com/show/64yoHwDdhBd0vik2sev6xs] Apple Podcasts: https://podcasts.apple.com/gh/podcast/shipping-excellence/id1794200466 [https://podcasts.apple.com/gh/podcast/shipping-excellence/id1794200466] YouTube: https://www.youtube.com/@shippingexcellence [https://www.youtube.com/@shippingexcellence] RSS: https://media.rss.com/shippingexcellence/feed.xml [https://media.rss.com/shippingexcellence/feed.xml]

No one cares about AI. They care about results.

AI doesn't fix your broken processes. It amplifies them. Jason Macht (Whitespace Solutions) joins Andrew Duncan to break down what actually changes — and what doesn't — when you build with AI. Chapters 00:00 Intro 00:40 Guest introduction 02:51 Vibe coding vs. engineering rigor 04:02 AI as amplifier — good and bad processes 04:39 Spec-driven development 06:57 The death of traditional agile 09:35 Agentic systems and guardrails 15:27 New business viability — the solopreneur opportunity 32:09 AI as thought partner and echo chamber risks 49:45 Closing Topics * Why AI amplifies good and bad engineering practices equally * Spec-driven development: engineering at the speed of thought * The difference between vibe coding and production-ready software * How solopreneurs can now build niche software businesses without VC * Agentic systems, guardrails, and what happens when AI has real-world access * Using AI as a strategic thought partner — and avoiding echo chambers * Why traditional agile is losing to fast betting cycles Guest Jason Macht — Founder, Whitespace Solutions * https://www.whitespacesolutions.ai/ [https://www.whitespacesolutions.ai/] * https://jasonmacht.com/ [https://jasonmacht.com/] * https://www.linkedin.com/in/jasonmacht/ [https://www.linkedin.com/in/jasonmacht/] Resources Mentioned * Cursor: https://cursor.com/ [https://cursor.com/] * Replit: https://replit.com/ [https://replit.com/] * Bolt: https://bolt.new/ [https://bolt.new/] Connect Andrew Duncan: https://www.linkedin.com/in/andrew-b-duncan/ [https://www.linkedin.com/in/andrew-b-duncan/] Vertice Labs: https://www.verticelabs.io [https://www.verticelabs.io] Subscribe Spotify: https://open.spotify.com/show/64yoHwDdhBd0vik2sev6xs [https://open.spotify.com/show/64yoHwDdhBd0vik2sev6xs] Apple Podcasts: https://podcasts.apple.com/gh/podcast/shipping-excellence/id1794200466 [https://podcasts.apple.com/gh/podcast/shipping-excellence/id1794200466] RSS: https://media.rss.com/shippingexcellence/feed.xml [https://media.rss.com/shippingexcellence/feed.xml] YouTube: https://www.youtube.com/@shippingexcellence [https://www.youtube.com/@shippingexcellence]

The Real Competitive Advantage in an AI World (It’s Not Tech)

In this episode of Shipping Excellence, Andrew Duncan sits down with Christopher Garvis (former AWS manager, worked on CDK) to unpack a core reality of the AI era: AI makes speed cheap, judgment priceless. They start with the practical: how experienced builders are using AI day to day, why “skills” (constrained context + tools) are emerging as the next standard beyond agents, and what happens when heavy integrations like MCP feel slow or token expensive. From there, the conversation shifts to the business implications. If AI compresses the time to ship, teams hit decision points faster, with less time to react. That makes judgment and ownership the highest leverage skills. They discuss why AI amplifies whatever process you already have, why “vibe coding” often hits a ceiling without scar tissue, and why the durable moat is still culture: integrity, follow through, and loving the customer by truly understanding them. They close on the human side of technology: authenticity in a world where it’s harder to know what’s real online, the importance of “third places,” and why remote teams need intentional in person time to build trust and momentum. Topics covered * Why AI makes speed cheap, and why judgment becomes the premium skill * Skills vs agents: constrained context, tool boundaries, and predictable workflows * Parallel AI workflows: running multiple streams across code and content * MCP vs lightweight tools: when a CLI beats a server integration * Emergent behavior: when the model bypasses tools for simpler paths * AGI and what “human” means beyond text: embodiment, emotion, and social bounds * AI amplifies process: discipline compounds quality, chaos compounds failure * Why ownership still matters: “AI isn’t responsible for outcomes, humans are” * Vibe coding, security gaps, and what breaks at real customer scale * The new moat: integrity, customer support, follow through, and reputation * Authenticity vs perfection in an AI saturated internet * Remote work done right: high bandwidth FaceTime, offsites, and trust currency * Community and “third places” as technology reshapes connection Key takeaways * AI accelerates execution, but it does not remove the need for judgment and ownership. * The faster you can ship, the faster weak assumptions and bad decisions compound. * Strong engineering discipline gets supercharged with AI; weak discipline gets exposed. * Technology is commoditizing, so culture and integrity become defensible advantage. About the guest Christopher Garvis is a veteran engineer and former AWS engineering leader who led the AWS CDK team. He focuses on practical AI workflows, tool constrained systems, and shipping with strong engineering fundamentals. About Shipping Excellence Shipping Excellence is a podcast about building exceptional software products: practical engineering, strong delivery, and the systems that help teams ship. Links * Vertice Labs: https://verticelabs.io [https://verticelabs.io] * Follow Andrew Duncan: https://www.linkedin.com/in/andrew-b-duncan/ [https://www.linkedin.com/in/andrew-b-duncan/] #ShippingExcellence #AINative #SoftwareEngineering #ProductDevelopment #Leadership #RemoteWork #Agents #Skills #MCP #AWS #StartupExecution

Why Most Software Vendors Fail Operators — And What Good Actually Looks Like

Most software projects drift — scope grows, timelines slip, and clients spend more time managing the vendor than shipping product. Robert Randolph, CEO of Executive Linguist Agency, hired Vertice Labs to build two apps. Both shipped on time. Both on budget. Here's what made it different. Robert runs a professional language services firm with a remote team of interpreters. He needed a HIPAA-compliant messaging platform and an internal iOS app — and he needed them fast. The experience became a useful lens on what actually makes software partnerships work: not charisma, not promises, just execution, clarity, and trust. Chapters 00:00 The first problem: getting an app in the App Store fast 02:50 Hesitations before starting — and why there weren't many 07:38 HIPAA-compliant interpreter messaging: the gap in the market 08:26 "You got us to market so fast I didn't have time to build a sales strategy" 10:01 What no-BS looks like in a pitch 12:20 $40K consulting with nothing to show for it 14:33 Executive Linguist: Vertice Labs' ideal client 17:29 How a Stripe security incident revealed the team's character 19:01 Speed and organization: the biggest differentiator 21:00 "You could have milked us — and you didn't" 24:20 Who should hire a software partner 31:15 What made this client relationship work 35:00 How watching the process improved their internal operations Topics • What makes a software partnership actually work • Clear scope of work as a trust foundation • Transparency, autonomy, and accountability on a dev team • Speed without sacrificing quality • What to look for in a software partner — and what disqualifies one • Building HIPAA-compliant apps for specialized industries Guest Robert Randolph, CEO — Executive Linguist Agency https://www.executivelinguist.com [https://www.executivelinguist.com] https://www.linkedin.com/in/robert-randolph-a86ab4111 [https://www.linkedin.com/in/robert-randolph-a86ab4111] Resources mentioned Linear — https://linear.app [https://linear.app] Stripe — https://stripe.com [https://stripe.com] TestFlight — https://developer.apple.com/testflight/ [https://developer.apple.com/testflight/] Connect Andrew Duncan — https://www.linkedin.com/in/andrew-b-duncan/ [https://www.linkedin.com/in/andrew-b-duncan/] Vertice Labs — https://www.verticelabs.io [https://www.verticelabs.io] Subscribe Spotify — https://open.spotify.com/show/64yoHwDdhBd0vik2sev6xs [https://open.spotify.com/show/64yoHwDdhBd0vik2sev6xs] Apple Podcasts — https://podcasts.apple.com/gh/podcast/shipping-excellence/id1794200466 [https://podcasts.apple.com/gh/podcast/shipping-excellence/id1794200466] YouTube — https://www.youtube.com/@shippingexcellence [https://www.youtube.com/@shippingexcellence] RSS Feed — https://media.rss.com/shippingexcellence/feed.xml [https://media.rss.com/shippingexcellence/feed.xml]