I
Build for ourselves
The first pass of any studio tool is built for a real page on the consumer site. The motion engine existed because a chart wheel needed to animate. The ornament library exists because reference pages needed corner marks and brass dividers. The OG template exists because 54 astrology URLs needed share cards. We start from a concrete job, never from a “maybe someone needs this.”
This keeps the surface area honest. No speculative configs. No half-built abstractions for a second use case that never arrives. The code shape is dictated by the problem it was written to solve.
II
Run it in production for weeks
Before anything graduates to the portfolio, it runs on insightsbyomkar.com under real traffic — not a demo site, not a staging URL, the live consumer surface. We watch it in real users’ devices: slow phones, old Safari, accessibility tools, screen readers, prefers-reduced-motion, low-bandwidth connections. A tool that only looks good on a MacBook on fibre is not finished.
That’s why the portfolio carries status badges: readymeans it’s been running long enough that we’ve seen every failure mode and fixed the ones worth fixing. Beta means stable but still sanding. Plannedmeans the engine exists but the API wrapper doesn’t.
III
Sell it as a real product
Graduation looks like: an API surface with docs, a pricing scope, and a pilot client on the other end. No public pricing menu — every pilot is scoped against the actual problem you’re bringing. That sounds slower than a SaaS signup, and it is. It’s also honest: you pay for the work that fits your shape, not for a tier designed to capture you at the wrong one.
Once enough pilots have scoped the same product to similar numbers, that’s when public pricing shows up. Not before.