Case title pending
Placeholder: e.g. "Manuscript rejected twice for structural issues; restructured and accepted in a Q1 journal after one revision round."
We write, edit and validate scientific communication — for researchers publishing their work, biotech teams building authority, and digital health products whose claims need to hold. The distance between what your data says and what your audience understands: that’s where we work.
Anyone can write with AI. Few can tell you whether the science behind your message is actually defensible.
The hummingbird’s green is not pigment — it is structure. A scientific text works the same way: the same data can dazzle or go unnoticed depending on how it is built.
"A statistically significant trend towards amelioration of the evaluated parameters was observed in the intervention cohort, which could potentially suggest therapeutic implications that may warrant further exploration in subsequent investigations."
This is what we do: close the gap between the data and its reading.
"Our AI optimizes your nervous system through vagus nerve activation."
From our science validation line — claims that survive a clinician’s reading.
A 30-minute call, no strings attached. We arrive with questions, not a catalogue.
Every deliverable starts from the same principle: the structure does the work. We don't sell words — we build and verify arguments that hold up in front of a reviewer, an investor or a feed algorithm.
From data to accepted manuscript. We work with your results and your team so the paper says exactly what the evidence allows — and reads the way it deserves.
Your team knows the science. We turn it into material that builds authority over time: for clients, for investors, for the field. Judgement, not just presence.
AI made building health products easy. It did not make the science behind them reliable. Before your app, landing page or pitch deck reaches users, investors or clinicians, we check whether the claims, the references and the evidence actually support what the product says.
| Product claim | Evidence | Risk |
|---|---|---|
| "Improves sleep quality" | Indirect studies; acceptable if qualified | Defensible |
| "Reduces anxiety" | Weak or extrapolated evidence | Needs rewording |
| "Detects early signs of depression" | Requires clinical validation | Medical device territory |
What this is not: a regulatory certification or legal assessment. We provide scientific, evidence-based review of product claims, messaging and supporting literature. For MDR or AI Act compliance, we'll tell you when it's time to bring in regulatory counsel — and we'll have done the groundwork they need.
Almost nothing does. Write to us with what you have — a dataset, an idea, a half-finished draft — and we'll tell you how we would approach it.
A sample of recent projects across our three lines. Much of what we write is published under our clients' names or protected by confidentiality agreements — these cases are anonymised, and we're happy to discuss relevant examples in detail on a call.
Placeholder: e.g. "Manuscript rejected twice for structural issues; restructured and accepted in a Q1 journal after one revision round."
Placeholder: e.g. "Response to reviewers for a contested methods critique — all major points resolved without additional experiments."
Placeholder: e.g. "White paper for a biotech platform company — now their most-requested sales document, 18 months on."
Placeholder: e.g. "Six months of LinkedIn ghostwriting for a healthtech founder — follower growth metric, inbound leads metric."
Placeholder: e.g. "Claim audit for a sleep app pre-launch — 14 claims mapped, 3 flagged as medical device territory, all reformulated."
Placeholder: e.g. "Evidence dossier for a wearable's HRV-based features, used in the company's Series A data room."
Papers where our contribution is acknowledged or where authors have agreed to be listed as references. List pending — entries will follow standard citation format.
Iridis was born from the conviction that well-written science is not a concession to popularisation: it is the standard to aspire to. We are a small team on purpose — every project goes through the same hands that scope it.
Latin genitive of iris: "of the iris", "of the rainbow". In scientific nomenclature, the genitive is the standard form. Three readings in a single word:
The hummingbird's structural colour. Not pigment: construction. Like a good text.
Precision of vision. Seeing clearly what the data says — and what it doesn't.
The bridge between science and those who need to understand it.
Complementary profiles: science, writing and strategy.
Short bio: background, years of experience, therapeutic areas or sectors of expertise. Two or three lines at most.
Short bio: background, years of experience, therapeutic areas or sectors of expertise. Two or three lines at most.
Short bio: background, years of experience, therapeutic areas or sectors of expertise. Two or three lines at most.
There are no account managers between you and the text. The person who understands your project is the one who writes it.
If the evidence doesn't support the claim, we'll tell you before a reviewer does. It's cheaper that way.
Scope defined per deliverable, revision rounds agreed up front, no surprises on the invoice.
We write directly in the target language. We don't translate: we write.
We reply within 48 working hours. If you have a draft, a dataset or a link, even better: we'll arrive at the call with our homework done.
Three things that speed up any quote: what you need (type of deliverable), by when, and for which audience. With that, we can give you a range in our first reply.