Show notes
In Part 1 of this conversation on Triangle Tweener Talks, we walked through how David built Offline over 13 years, three companies, multiple pivots, and a subscription model that finally worked.In Part 2, we shift gears. This episode is about leverage. David breaks down how he went from running a ~34-person team to operating a multi-city business with ~2.5 people, not by cutting corners, but by rebuilding the company around AI-first systems, orchestration, and context-aware automation.This is not “AI news.” This is how a founder is actually using AI day to day.Highlights from Part 2AI adoption started as a “copywriting intern,” not a silver bulletThe biggest fork: people who learned how to talk to models vs. people who gave upThe real unlock came when David built his first full-stack app himselfTools like Cursor and Cloud Code collapsed the barrier between idea and executionOffline stopped adding features to a monolith, and started building around itn8n became the orchestration layer that glued everything togetherMost business problems don’t need apps, they need glue codeAI SDRs fail today because context is fragmented and CRMs are a messThe right approach is decomposing SDR work into atomic stepsContext windows are the real constraint, not intelligenceDavid now runs a personal “Chief of Staff” GitHub repo with 70–80 skillsThe company itself is slowly becoming a file system agents can read and write fromIf you’re thinking about automation, agents, or headcount, this one will change how you think about all three.Where to Find David Shaner:LinkedIn: https://www.linkedin.com/in/davidshaner/Offline Media: https://www.letsgetoffline.com/Where to Find Scot Wingo: LinkedIn: https://www.linkedin.com/in/thescotwingo/ Tweener Times: https://www.tweenertimes.com/ X: https://x.com/scotwingoTimestamps00:00–02:30 – Why Part 2 matters02:30–06:00 – Phase 1: AI as a copywriting intern06:00–10:00 – Phase 2: coding, Cursor, and the “hit by a bus” moment10:00–13:00 – Why founders need to build something themselves13:00–17:00 – Orchestration layers and why n8n won17:00–21:00 – Why AI SDRs mostly don’t work yet21:00–24:30 – Context windows, atomic steps, and agent design24:30–27:30 – Debugging workflows like a human27:30–31:30 – How Offline reduced headcount without losing velocity31:30–36:00 – Self-service platforms for restaurants and events36:00–39:30 – Cloud Code, skill files, and the “singularity” moment39:30–44:00 – The Chief of Staff repo and what comes next--- This episode of Triangle Tweener Talks is hosted by Scot Wingo, presented and produced by Triangle Tweener Fund, with creative assets and design support from Walk West. We couldn’t share posts like this without our amazing sponsors: Gold Sponsors: Balentine: https://www.balentine.com/triangle-en... EisnerAmpner: https://www.eisneramper.com Robinson Bradshaw: https://www.robinsonbradshaw.com Silver Sponsors: Automated Consulting Group: https://automated.co Bank of America: https://business.bofa.com/en-us/conte... 2025 Sponsors: Extensis HR: http://www.extensishr.com/------Triangle Tweener Talks is sponsored by:Atomic Object: https://atomicobject.com/
