No-CodeAssetFiltering

I knew from experience that the inherited designs would fail in production, but the Product Owner wanted to ship them as-is to hit a deadline. When my critique was dismissed as just a designer's opinion, I proposed a bet: we run a rapid internal test. If users succeed, we build the spec; if they fail, we fix it.

The test surfaced failures so severe that building the spec became a bigger risk than delaying it. Faced with hard evidence rather than a designer's opinion, the PO changed course and funded two weeks of redesign before engineering wrote a single line.

Because our users were seasoned CCIEs, I prototyped a raw SQL editor as an alternative to the inherited no-code builder. It's tempting to build power features for power users, but complexity is a tax the other 99% pay.

I used my judgment to scrap the SQL exploration. I kept the no-code direction, knowing we could add advanced operators later to close the power gap without sacrificing baseline usability.

The project was scoped as a seven-month waterfall plan. The organization assumed the feature had to be completely finished before it could ship. But the existing failures were too glaring to leave in production for another half-year.

I negotiated with PM and engineering to drop the waterfall plan, and personally led the sprint planning to carve out just the critical fixes. Trading the massive rollout for an MVP cut shipping from seven months to two.

Resolving three years of usability debt

Unprompted feedback in the internal Cisco forums confirmed we had finally resolved the usability issues that had generated complaints for three years.