Citation manipulation leaves a trail. PeerTrust surfaces it.
PEERTRUST, INC.
PeerTrust turns black-box suspicion into defensible evidence — transparent metrics, preserved source snapshots, and reviewer-grade dossiers that stand up to appeals — for universities, publishers, funders, and integrity committees.
From idea to company.
Started with a Collaboration between ScienceWerx, Inc and KNOWDYN, UK
ScienceWerx, Inc and KNOWDYN, UK, worked together to develop a "Know Your Scientist" that started with a book, KNOW YOUR SCIENTIST
KNOWDYN, UK Built the Algorithm
Built and tested the Algorithm
In the process of forming a Delaware C-Corp
Expected to form a Delaware C-Corp
ScienceWerx (Gx Lab Task Force Built the Platform)
Built a working Platform and opened it to the public for Testing
Citation gaming has industrialized. Citation mills sell citations openly, researchers have demonstrated buying citations for a fictional author, and coordinated cartels quietly distort rankings, hiring, and funding decisions. The tools built to catch this behaviour mostly answer with a proprietary risk score — a number that can start an investigation but cannot survive one. When a researcher appeals and a committee asks how the figure was calculated, "our vendor's algorithm flagged it" is not an answer.
PeerTrust was built on a different premise: detection isn't the hard part anymore — defensibility is. Instead of a black-box score, PeerTrust builds the case file. Working entirely from open scholarly data (OpenAlex and ORCID), it computes three transparent, deterministic signals: Collaboration Isolation, which measures how inward-facing an author's citation neighbourhood is; Reciprocal-Source Pacts, which surfaces bidirectional citation exchanges between the same sources; and Publication Bursts, which flags citation-velocity spikes against an author's own historical baseline. Every formula is published, and every value is reproducible from stored source snapshots.
Each investigation follows a six-step forensic pipeline — identify the researcher, harvest the publication corpus, build the citation network, measure reciprocity, measure collaboration isolation, and generate the dossier. The result is a hashed, reproducible evidence package with a complete chain of custody: interactive and print-ready reports, a per-signal risk radar, an evidence graph, a chronological timeline, full citation sources, network metrics, and the raw data behind them all.
Just as important is what PeerTrust refuses to do. It presents anomaly indicators for human-led review — never findings of misconduct, and never automated verdicts. It clears as often as it flags: when no indicator crosses its threshold, the researcher is cleared and nothing is escalated; when one does, the case is handed to a human reviewer. AI supports investigators, but expert judgement and institutional oversight make every decision.
Built by ScienceWerx and KNOWDYN, PeerTrust serves the institutions responsible for research integrity — university research offices, integrity committees, funding organizations, and publishers. It is live today as a free public beta: enter an ORCID and run a forensic investigation, no login required.