Emanuele Di Pietro did not set out to build a polished enterprise devtool. The MRR Story case describes a 22-year-old computer-science student in Naples who was already living inside local AI coding agents. The frustration was simple: when Codex was running on his Mac, leaving the desk meant losing the ability to steer the session, answer prompts, push commits, or recover from a stuck run.

His first version was more ambitious than the final wedge. He had been exploring Phodex, a cloud-style remote setup with VPS and Docker pieces. Then rumors surfaced that OpenAI might ship an official mobile/remote direction. Instead of competing with a heavy cloud product, Emanuele narrowed the promise: ship a local-first iPhone remote for Codex before the official app arrived. Around midnight on March 8, 2026, he posted the challenge publicly, used Codex itself to help build, and had the first working version by about 4 AM.

Remodex's durable hook is trust. The live product page and App Store listing both frame it as a local-first Mac bridge plus iOS app: Codex, git, and workspace actions stay on the user's Mac while the phone sends instructions through an end-to-end encrypted session. The GitHub repo reinforces that boundary with open-source code, self-hosting notes, and a bridge architecture rather than a black-box cloud IDE.

Distribution came from proof, not paid ads. Emanuele launched in public, put the app on the App Store, and made the bridge open source so technical users could inspect the trust model. MacStories later reviewed Remodex as the best Codex remote client for iOS until an official OpenAI app exists, highlighting the same local Codex setup, App Store availability, and subscription pricing. MRR Story reports roughly $10,000 revenue within about 30 days. That is strong proof of paid demand, while still being a first-month sales figure rather than proof of steady recurring revenue.