Build in public
Product decisions, architecture tradeoffs, and practical lessons from building observable and retryable AI content workflows.
Written and maintained by SymphonyIceAttack. Last updated .
A practical framework for indie founders to evaluate AI product ideas across instinct, consumption, creation, and work before committing to development.
Read article
A first-person engineering case study on rebuilding an AI content generation workflow using executable plans, independent units, and asynchronous Gemini Batch jobs.
Read articleThis blog explores what it takes to turn an AI product idea into something useful, understandable, and sustainable. It brings together first-person build notes, product experiments, validation frameworks, design decisions, practical tutorials, and lessons from operating WriteGeniuses. Individual articles may focus on a technical problem, a customer need, a go-to-market question, or a broader shift in how people create with AI.
The scope is intentionally open. Topics can range from architecture, reliability, and AI workflows to product strategy, user research, content quality, SEO, distribution, pricing, and indie product development. Category pages currently organize articles into engineering and product sections without limiting what the blog may cover next.
Articles separate direct observation, interpretation, and external evidence wherever that distinction matters. A case study may begin with product behavior or user feedback; a tutorial may rely on platform documentation; and a product essay may combine market signals with first-person experience. The aim is to make the reasoning visible so readers can decide which conclusions apply to their own context.
Sources vary with the subject. For technical articles, primary references may include the official Google Gemini API documentation, Convex documentation, and TanStack Start documentation. Other articles may cite research, public datasets, product documentation, or original experiments. Sources establish the facts; the articles contribute context, decisions, and lessons learned.
