Long-form pieces in SaaS, fintech, healthtech, AI/ML, devtools, infrastructure, compliance — written for the reader who'd notice if the bot wrote it. Voice-matched to your brand. Citations that hold up. SEO-aware without SEO-stuffed.
Your audience reads. They notice when ChatGPT wrote it. So do their colleagues.
Generic AI content has a tell. Three-sentence paragraphs that say nothing specific. Bullet lists where every bullet starts the same way. Buzzwords without anchors. Citations that go to nowhere. The reader's reflex kicks in by the second paragraph and the piece is dead.
And your domain — fintech, AI/ML, devtools, healthtech, compliance — is exactly the audience that does notice. They read this stuff for a living.
Domain writing is different work. AI does the structural lift — research synthesis, fact-checking, headline variants, SEO mapping. Humans handle the parts that AI can't fake: voice, specificity, the named systems that anchor a claim, the doctrinal closer that lands.
The output reads like a well-researched generalist who reads carefully. Not a PhD in your subfield. Not a chatbot pretending to be one. Where the topic demands true expert authorship, we say so upfront.
One human-gated intake. Two production stages. One human-gated review. Voice matching happens during stage 2 — not as a polish pass after the fact.
Topic brief, 2–3 of your existing pieces for voice, target audience, keywords or angle. 15-minute call if the brief is dense or domain-specialized.
AI-assisted research pass against current primary sources. Draft against your voice samples — not after. Named claims fact-checked before they land in the draft.
Voice match tightened by hand. SEO integrated into headlines, intros, sub-headers — not sprinkled into paragraphs. Citations linked. Headline variants drafted.
You react. We tune. One to three revision passes by tier. Final delivery in your preferred format with reference list, meta description, and social snippet.
If a paragraph reads as machine output on review, we rewrite it before delivery. The brand voice gate from company_ethos/08_BRAND_VOICE_GUIDELINES.md applies on every piece: no AI-hedging filler, named systems where they earn a sentence, em-dashes for force, declarative posture, sections end on the strongest line.
And we don't fabricate. Made-up statistics, invented studies, hallucinated citations — none of it lands in your piece. If we can't anchor a claim, we don't make it.
Pick by length and depth of voice-match. Single short piece at Entry. Standard flagship at Standard. Series or 5K-word flagship at Premium.
Most AI content fails for a single reason: the voice match happens after the draft, as a polish pass. The result is generic prose with a few brand words sprinkled on top. Domain readers spot it instantly.
We do it the opposite way. You send 2–3 pieces by your team at intake — ideally by the same writer if you want voice continuity. We read them before drafting. Sentence rhythm, paragraph length, the way you handle technical claims, the doctrinal closers your brand uses. That becomes the spine of the new piece. AI does the structural research lift. Humans do the voice and the specificity.
For specialized claims — clinical accuracy, regulatory specifics, code-level precision — we flag them in the draft for your subject-matter expert to verify before publication. If the topic demands true expert authorship, we say so upfront and recommend the interview-based piece instead.
Submit the form. Scoping call within 24 hours if the brief is dense. Most pieces get a written scoping confirmation back same-day.