The AI Stack We Actually Use for Prototyping, Strategy, and Personal OS

Peter demonstrates using a Claude project with a one-page strategy template and research to draft a memo for Google AI Studio. Tal says the first draft is vague and recommends starting with the team’s north star, success measures, and longer-term plan. The group refines the idea around winning in prototyping first, with a longer-term path from prototype to production.

Peter uses Google AI Studio to reproduce an existing interface and add a template feature, creating a shareable prototype teammates can iterate on. He says he tested a prototype with users before the designer’s Figma mocks were ready. The group suggests prototypes can help teams explore ideas and communicate them, while documents remain useful for clear thinking and edge cases.

Tal shows an Obsidian Kanban board backed by Markdown files that he can also open in Cursor. He asks AI to review initiative priorities and rollout plans, then discusses the recommendations before applying changes. He prefers reviewing diffs and controlling edits, especially when working with plain-language planning files.

Aman demonstrates a personal operating system using Claude Code, Obsidian, Markdown files, and skills for tasks such as morning planning. The workflow can check meeting transcripts, tasks, and goals, then suggest a daily focus; a local script accesses Granola notes. Aman says adding personal goals and work context helps the agent align suggestions with what matters to him.

Peter says AI helps him draft memos and PRDs, prototype with teams and users, stay focused across weeks, and handle routine emails in his writing style. Tal uses AI to think through initiative priorities and rollout stages. The hosts describe their customized workflows as personal AI products they keep improving, and emphasize that much of the setup is text files and context.