Open repositories cannot take it
Zenodo, Figshare and Dryad are excellent, and open data belongs in them. But they publish what you deposit. Data with consent conditions, ethics constraints or commercial sensitivity has nowhere equivalent to go.
Built in Malaysia, for this region
Share the data you cannot make open.
Some research data carries consent conditions, ethics constraints, or an agreement that says who may use it and for what. Emperix is built for that data: you set the terms, approve each request yourself, and can withdraw access at any time.
Open data belongs in an open repository. This is for everything else.
EXAMPLE RECORD
Household fever survey, 2023
- ACCESS
- By request — you approve each applicant
- CONCERNS PEOPLE
- Yes
- LAWFUL BASIS
- Informed consent obtained
- DE-IDENTIFICATION
- Pseudonymised (key held separately)
- PERMITTED USE
- Academic research only
- ETHICS REFERENCE
- ACCESS TERM
- 1 year
ACCESS LOG
Postdoctoral researcher, public university
Purpose: re-analysis of fever duration by age band
APPROVED 1 year · expires 14 Mar 2027
Analyst, private company
Purpose: product development
DECLINED outside permitted use
PhD candidate, public university
WITHDRAWN by the depositor
Example record. It shows the fields a depositor completes and the log that builds up behind them. It is not a real dataset, and the people in it are not real.
The first question
Is this not what Zenodo does?
No, and if your data can be open you should use Zenodo — it is free, well funded and permanent. Zenodo publishes what you deposit. Emperix is for data you cannot publish: consented human data, restricted industrial data, anything governed by an agreement. The two solve different problems and we would rather you used the right one.
Where we are
Emperix is new. The platform is built and running, and we would rather say exactly what that means than imply more.
The exchange is also new enough to be small. We are not going to dress that up with numbers.
The problem
Restricted data still moves by email
Most research data that matters is not open, and cannot be. It was collected under consent that limits its reuse, or it identifies people, or it belongs partly to somebody else. So it travels as attachments, governed by agreements nobody can later find.
So it moves as attachments
A request arrives by email. An agreement is signed as a PDF and filed somewhere. A file is sent, or a link is shared. Six months later nobody can say which version was sent, or on what terms.
And nobody can prove anything
When an ethics committee asks who accessed the data, for what purpose, and under which agreement, the honest answer is usually that no one kept a record. That is a difficult position for the researcher who shared it in good faith.
How it works
Four steps, and you keep control at every one
- 01
Deposit
Terms recorded, file screened
- 02
Describe
Description public, data withheld
- 03
Decide
Each request judged by you
- 04
Retain control
Terms expire, access withdrawn
The life of a grant
Access is a state, not a one-way door
A grant moves through states you can see and change. Nothing about it is permanent unless you decide it should be. Expiry and withdrawal are the two endings, and both are yours to choose.
Applied for
A named person states a purpose and an institution.
Approved
You grant it, for a term you choose.
Then either
Expiring
The term runs down. No renewal is automatic.
Withdrawn
The next download attempt fails, immediately.
For institutions
An answer when someone asks who had the data
Research offices, ethics committees and data stewards carry the risk when restricted data is shared. Emperix is built so that risk is documented rather than assumed: every application, every decision, every withdrawal, with the purpose stated and the institution named.
The record cannot be edited or deleted, including by administrators. An audit trail that the wrong person can quietly amend is not an audit trail.
Questions
Where the answer is no, it says no
- Do you mint DOIs?
- Not yet. DOIs require membership of a registration agency and we have not taken that step. Datasets are versioned and have stable addresses, so the groundwork is there, but we are not going to describe anything as citable until it genuinely is.
- What does it cost?
- Nothing at the moment. There is no payment system in Emperix, so nothing can be charged even if we wanted to. Pricing will be settled openly and in advance of any change; we are not going to charge people who deposited on the understanding that it was free.
- Who can see my data?
- For controlled datasets: you, whoever you have approved, and nobody else. The description and terms are public so people can tell whether to apply. The data is stored where it cannot be read directly — every download is a short-lived link issued only after your grant has been checked. Emperix staff do not browse deposited data.
- What if a participant withdraws consent?
- Withdraw the access. It takes effect at the next download attempt rather than at some later sync, and the file stops being reachable. Anyone who already downloaded a copy has a copy — no platform can recall that, and any platform claiming otherwise is misleading you. What Emperix can tell you is exactly who that was, when, and what they agreed to.