Real vs. Ideal AI Workflows: A Deep Dive with David Shaner and Taylor Cotner

AI demos are clean. Real AI workflows are... less clean.

In episode 2 of [REDACTED], David Shaner and Taylor Cotner get into what it actually looks like to use AI inside a real operating business. They talk through:
- Rebuilding lead-gen workflows after a major team compression
- Why HubSpot cleanup is much harder than it sounds
- The limits of n8n-style automation
- Building multi-agent systems that can research, reason, validate, and ask for help
- The joy/pain of keeping AI agents alive like Tamagotchis
- Using Claude Code to generate a real shareholder update from financials, GitHub commits, payroll, board notes, and past emails

AI is not magic. It is infrastructure, context, iteration, and a lot of “why is this broken now?” When it works, it can turn scattered business context into something genuinely useful.

Watch/listen to the full episode here:

🎧 Listen here:
📺 Watch the episode:
👉 Read about it on NC Tweener Talks:
#AI #Startups #Automation #ClaudeCode #AgenticAI #BusinessOperations