Anish Dahiya
← Back to all writing

Apr 21, 2025 · 9 min

Building a Creator Flywheel as a Full-Time Data Scientist

The frameworks, templates, and automations that let me ship high-leverage content while leading teams.

I run a content flywheel while leading data teams by batching energy, codifying prompts, and automating the dull parts. Here's the operating system.

Define the thesis and voice

Every quarter I pick three themes: applied AI, systems for builders, and creator business mechanics. Anything outside gets parked. This protects creative focus.

Voice is scripted in a 'tone board'—short, direct sentences with concrete metrics. It keeps GPT assistants and collaborators aligned with how I speak.

Capture > polish

Daily capture sessions happen in mem.ai. I drop voice notes, screenshots, and code snippets tagged by theme. Once a week I promote the most resonant notes into outlines.

Automation handles formatting: a custom script turns outline headings into LinkedIn carousels, newsletter drafts, and video shot lists.

Ship consistently without burning out

Content drops run on a kanban: capture → outline → draft → publish → repurpose. Each state has templates, so collaborators can step in.

A feedback ritual closes the loop: I log hook performance, watch time, and replies inside Airtable to decide what to double down on next week.

Key takeaways

  • Focus on three themes per quarter
  • Automate formatting so creativity stays high
  • Track performance like a product funnel

A flywheel is just a series of small systems. Once capture, editing, and distribution run on rails, you can show up as a creator without sacrificing your day job.