We rewrite AI text so it reads like a human wrote it.

Inkognito started in late 2024 as a side project. One of us was drafting a research paper and using ChatGPT to accelerate the boring parts — the citations, the transitions, the abstract. The writing was technically correct, but the department's detection tool flagged it anyway. Nothing was plagiarized. Nothing was invented. The work was ours. But the rhythm of the sentences gave it away.

That's when we realized the problem wasn't the AI. The problem was that current detectors don't look at meaning — they look at patterns. Perfect grammar. Predictable transitions. Sentences that all breathe in the same rhythm. Fix the rhythm, and you fix the score. So we spent the next six months studying how detectors actually work, reverse-engineering the signals they flag, and building a rewriter that fights uniformity instead of hiding it.

The first public version of Inkognito shipped in early 2025. Today, thousands of writers use it every week to reclaim their voice from the "AI-generated" label.

Most humanizers swap synonyms. That's a fingerprint. If you replace every "utilize" with "use," a detector notices the pattern just as easily as it noticed the original. We take a different approach.

Inkognito's rewriting engine targets the structural signals detectors depend on: sentence-length uniformity, parallel clause construction, overly polished transitions, and the tell-tale absence of the small, messy artifacts that real human writers leave behind. Instead of masking these signals, we introduce controlled human-like variance — mid-sentence reconsiderations, near-repeated phrases, sentences that abandon one idea for a better one halfway through.

Every rewrite runs through a heuristic scorer before it's returned to you. If a passage still reads too uniform, our engine loops back and rewrites just the flagged sections — not the whole document. That's why Inkognito holds its score across long documents where other tools fall apart after the first few paragraphs.

Inkognito is built and maintained by a small team based in Nairobi, Kenya, with contributors in London and Toronto. We are writers, engineers, and one very opinionated linguist who insists the Oxford comma is non-negotiable.

We are independent. No investors, no ad-tech tracking scripts, no data brokers. The business is funded entirely by the writers who pay for it — which means we answer to them, not to a growth chart.

We don't guarantee 100% bypass on every detector, forever. Detection tools update, and so do we — but any tool that promises permanent undetectability is selling you a fantasy. What we do promise: transparent scores, honest results, and a rewriter that keeps improving as the detectors do.

Inkognito is a writing tool, not a substitute for doing the work. It's most useful when you already have something to say and just need it to sound like you.

Questions, feedback, feature requests, or a strongly-worded complaint about the Oxford comma — we want to hear it. Reach us any time.