For research groups and research offices
Research infrastructure, not a software purchase.
Ethical AI practice, metrics you can actually read, and a portfolio view of what your group has produced. Built by people who submit to journals themselves — which is why none of it is a dashboard.
Four engagements
What we actually do.
Each one is described by the decision it improves rather than by what it is made of, because what it is made of is mostly conversations and documents.
Ethical AI workflows for research groups
What your group may use a model for, what it must never touch, and how each of those is written into a disclosure an editor accepts. Delivered as a working practice your postdocs actually follow rather than a policy nobody reads — because the alternative is that everybody uses these tools anyway and nobody declares it.
Decides: What your group can defend in a submission
Research-metrics literacy
How to read the indicators your institution already reports on, what each one does and does not measure, and where a number is being asked to carry a decision it cannot support. We are not selling you a ranking; we are making the ones you are given legible.
Decides: What your committee should and should not conclude
Publication-portfolio review
A department or a group read as a body of work rather than a list: where the output is concentrated, where it is thin, which venues are doing the work and which collaborations are load-bearing. Written for a head of department deciding where next year goes.
Decides: Where the next three years of effort should go
AI-use policy drafting
A policy your research office can publish and your researchers can follow, that survives contact with a journal, a funder and a research-integrity officer. Drafted with the people who will have to enforce it, and short enough to be read.
Decides: What your institution says when somebody asks
Where the line is
Four things this is not.
This is the part of the site most easily mistaken for a vendor pitch, so here is what we are not doing.
- We do not sell a dashboard, a licence, or a platform your IT team then has to support.
- We do not rank researchers, score departments, or produce a number a promotion case can rest on.
- We do not run your systems, integrate your CRIS, or take on anything that belongs to an IT function.
- We do not draft a policy and leave. The people who enforce it are in the room while it is written.
If what you need is a system integrated or a platform supported, we will tell you on the first call and point you at somebody who does that properly. Turning work away is cheaper than doing the wrong work, for both of us.
Non-negotiables
Four commitments we publish, so you can hold us to them
Human-Led Certification
Every strategic decision is reviewed and signed by a PhD expert — never solely by an automated system.
Ethical AI Disclosure
We disclose every tool used in your diagnostic. No hidden automation.
Data Sovereignty
Your intellectual property remains yours. We never use your work to train a model.
No Ghostwriting, Ever
We nurture your voice. We never substitute your ideas with ours — and we will decline work that requests us to.
Start here
Book an institutional call.
Thirty minutes on what you are trying to decide and whether any of this helps. Scoped and quoted afterwards, invoiced against a purchase order.
Book an institutional call
30 minutes. For deans, teaching centres, graduate schools and research offices. Scope, cohort size, delivery mode and what your finance office will need from us.
The AI tools used on a client’s work are named in their report, and the commitments above are published rather than described. Read the integrity framework before you decide whether to talk to us about AI policy — it is the fastest way to tell whether we mean it.
Quick answers
Objections, handled
Is this IT consulting?
No, and the distinction matters more than it sounds. Every piece of this work is about an academic decision — what a panel will ask, what a promotion committee should conclude, what a research-integrity officer will accept. We do not run systems, integrate platforms or support software. If your problem is a CRIS migration, we are the wrong people and will say so on the first call.
Will you tell us our h-index went up?
No. We will help you read the indicators you are already given, including the parts where a number is being asked to carry a decision it cannot support. Selling a department a metric that flatters it is easy and worthless; explaining what a metric actually measures is harder and is the thing worth paying for.
What do you need from us to start?
A conversation, and then usually less than people expect: published outputs are public, policies are documents, and the AI work starts from what your group is already doing rather than from an audit of everybody. We do not need identifiable data about individual researchers, and we will push back if somebody offers it.
Who does the work?
The founder, Dr Xu-Priour: a CNU-certified lecturer, published in A- and B-ranked journals. That is deliberate: policy written by somebody who has not submitted to a journal in a decade is policy that fails on contact with a real submission.
What does it cost?
It is scoped and quoted rather than listed, because a research office with one policy question and a group rebuilding a portfolio review are not the same engagement. You get a written scope and a figure together, before anything is signed, and it is invoiced against a purchase order like everything else institutional.