Principle Comparison
Compare how different speakers approach the same concepts. Select a topic to see various expert perspectives.
Quick Comparison Matrix
Click any dot to jump to that speaker's approach for the selected concept.
| Speaker | Agent Architecture(viewing) | Debugging Approach | Model Selection | Context Management | Automation Philosophy | Testing Strategy | Creative Philosophy | Learning Approach | Human-AI Collaboration | Prompting Strategy | Tool Integration |
|---|---|---|---|---|---|---|---|---|---|---|---|
DS Dan Shipper | — | — | — | — | — | ||||||
BT Ben Tossell | — | — | — | — | — | ||||||
AM Ashe Magalhaes | — | — | — | — | — | — | — | ||||
RC Ryan Carson | — | — | — | — | — | ||||||
NQ Natalia Quintero | — | — | — | — | — | — | — | — | — | — | |
KP Katie Parrott | — | — | — | — | — | — | — | — | — | — | |
NE Nat Eliason | — | — | — | — | |||||||
TH Tina He | — | — | — | — | — | — | — | ||||
PD Paula Dozsa | — | — | — | — | — | — | |||||
CH CJ Hess | — | — | — | — | — | — | — | ||||
LK Logan Kilpatrick | — | — | — | — | — | — | — | — | — | ||
GL Geoffrey Litt | — | — | — | — | — | ||||||
KR Kevin Rose | — | — | — | — | — | — | — | — | — | ||
TS Thariq Shihipar | — | — | — | — | — | — | — | — | — | — | |
NN Naveen Naidu | — | — | — | — | — | — | — | — | — | — | |
YP Yash Poojary | — | — | — | — | — | — | — | — | — | — | |
BB Brooker Belcourt | — | — | — | — | — | — | — | — | — | — |
Detailed Approaches
Explore how different builders approach each concept
How different builders structure their AI agents and orchestration systems
Dan Shipper
CEO & Co-founder, Every
Agent-native design with parity principle. Codebases should be structured for AI comprehension with granular, skill-based organization. Each component should have clear boundaries that both humans and AI can navigate.
"Design your codebase with parity—the same mental model for humans and AI."
Nat Eliason
Author & Founder, Independent
Multi-model orchestration using Opus as the brain and Codex for execution. A persistent Mac Mini runs Clawdbot 24/7, coordinating tasks via Telegram and managing long-running workflows across multiple projects.
"Use the right model for the right job—Opus thinks, Codex executes."
Ryan Carson
Founder, Untangle
Ralph Loop with structured JSON files for task tracking. The system reads tasks, executes them autonomously, marks completion, and continues until all work is done. Progress is tracked in external files.
"Let the AI work through a task list while you sleep."
Ashe Magalhaes
Founder, Harf
Personal AI suite with a central hub architecture. One coordinator manages multiple specialized agent contexts (coding, writing, research), with Slack integration for seamless communication across domains.
Geoffrey Litt
Engineer, Notion
Notion as a Kanban board for Claude Code tasks. Use existing productivity tools as the orchestration layer, with agents reading from and writing to shared databases. Bridges between external agents and app-native agents.
"It's definitely not going to be like each app has its own agent and they don't talk to each other."
CJ Hess
Engineer, 10X
MCP-based architecture that teaches Claude every framework. Build Model Context Protocol servers that provide agents with up-to-date documentation and tooling knowledge on demand.