Kdemic began with a problem its founder had as a researcher, not with a market study: language tools that improved a manuscript's English while quietly removing the person who wrote it.
The problem it was built for
A researcher writing in English as a second language is repeatedly asked to make a trade. Submit in your own words and risk a desk rejection on language. Hand the manuscript to an editing service or a general-purpose model and get back something publishable that no longer argues the way you argue. Neither outcome is acceptable, and the choice itself is an artefact of how academic publishing is organised rather than a fact about the researcher's ability.
A second problem sits underneath the first. Text that reads as authoritative can quietly outrun what its citations support. When nothing connects a sentence to the source it rests on, responsibility for that sentence becomes hard to locate, and the people who suffer are the reviewers, the readers, and eventually the author.
What Kdemic is
A supervised academic writing environment. It drafts against how a researcher actually writes rather than against a general model register, checks empirical claims against retrieved literature before they reach the researcher, and records each decision in a record the researcher can export.
It is not a paraphrasing service, it is not offered as a way to influence how any detection system classifies a manuscript, and it is not designed to produce a finished manuscript without the author in the loop. The commitments that govern what it may do are set out in our Responsible AI principles.
Where it comes from
Kdemic follows from public work on research integrity rather than sitting next to it. Its founder developed the open screening measures published at KelleniIndex.com — ARI², JPI² and PPGDI — with methods stated openly, code and datasets deposited, and a governing principle carried over into this project: a flag is a reason to look more closely, not a verdict.
The clinical work matters here too. A protocol that is used on patients has to be written down in a way other clinicians can check, argue with and reproduce; the same is true of a screening index, and of a manuscript. Kdemic is built on the assumption that the writing is part of the accountability, not a presentation layer added afterwards.
That work shaped the design. A system that asks researchers to be accountable for their claims has to be accountable for its own, which is why the record is exportable, the limits are published, and the things Kdemic will not build are written down where they can be quoted back at us.
Who is building it
Mina T. Kelleni, MD, PhD — founder and inventor. Lecturer of Pharmacology (equivalent to Assistant Professor) at the College of Medicine, Minia University, Egypt, and Research Fellow at INTI International University, Malaysia. Listed among the Stanford/Elsevier top 2% most-cited scientists for four consecutive years, 2022 to 2025.
He is the originator of Kelleni’s protocol, an immunomodulatory regimen for the early management of COVID-19 and other respiratory viral infections. It was first proposed in Pharmacological Research in 2020, and later published under its own name in Inflammopharmacology (2022) and in The Journal of Infection (2023), where it was reported in real-world use across paediatric, adult and pregnant patients. He founded Egypt’s first Pharmacology Clinic, and a Post-COVID and tinnitus clinic.
As an inventor he holds international copyright on the topical use of berberine in male sexual disorders, commercialised as PLAZMID topical gel and approved by the Saudi Food and Drug Authority as a cosmetic (record CN-2023-373373), distributed through pharmacies across the Kingdom. He also holds international copyright on a novel combined valproic acid and topiramate neurological formulation, with a patent application filed in Saudi Arabia; that compound has not yet entered clinical trials. He developed the research-integrity tools at KelleniIndex.com.
Kdemic is currently a solo project supported by invited researchers. ORCID 0000-0001-6290-6025.
Stage, stated honestly
Kdemic runs today as a working private build used by the founder and a small number of invited researchers. It is not a hosted public product, it is not sold, and there is no waiting list that converts to a subscription. The next milestones are a controlled public MVP and supervised pilots with researchers and institutions.
The underlying computer-implemented methodology is the subject of U.S. Provisional Patent Application No. 64/079,374, filed 1 June 2026. Implementation detail is shared only under appropriate confidentiality and partnership arrangements.
Working with us
We would like to hear from a technical co-founder or engineering partner, from universities, publishers and editing providers willing to run a supervised pilot, from advisers in responsible AI and publication ethics, and from accelerators and pre-seed investors working on research infrastructure. The contact form reaches the founder directly.