Started with four projects and no way to keep them straight

I was running four different projects at the same time. Each one needed social content across multiple platforms. I tried doing it manually for a while, and it was not sustainable. So I started using AI tools to generate the posts.

The output was fast. It was also wrong. Not wrong in an obvious way. Wrong in the way that you read it back and think, "I would never say it like that." A phrase here, a sentence structure there, a level of formality that did not match. Multiply that across four projects, up to 7 different platforms, and up to three posts per week each had a different voice, and every generation needed 15-20 minutes of editing before I could post it, I did not have this kind of time.

The problem was not the AI. The problem was that I had no way to tell the AI what my rules actually were. I could describe the tone in a prompt, but I could not say "never use these 25 phrases" or "this project uses contractions, that one does not" or "LinkedIn should sound different from Reddit." Every tool I tried treated voice as a vague prompt-level adjustment. None of them let me define hard rules and enforce them, and even when I tried to add voice rules in the setting, they never stuck and I was constantly correcting them.

So I built an app. SplitPost lets you define your voice as explicit constraints: banned phrases, sentence patterns, vocabulary rules, formality, per-platform notes. A two-pass system generates platform-native content and then validates every line against those rules. Violations get rewritten before you see the output and are shown in real time.

It started because I needed it for my own projects. Now it is live at splitpost.io. Free tier gives you 3 generations to test it.