Research community

Research skill, evaluated by researchers

Vocaid is building an evaluation community for working scientists: structured review of research ability — not citation counts — scored against published rubrics by evaluators whose own calibration is on record.

The founding evaluator cohort is now forming across eight disciplines.

Apply to evaluateRead the partner page

What this community is

Most of what makes a good researcher is invisible to the systems that credential one. Citation counts lag by years, h-indexes reward field size as much as insight, and hiring committees outside academia rarely have the expertise to read a publication record critically. We are building a different instrument: a community of working researchers who evaluate evidence of research skill directly, against standards that are published for anyone to read.

The unit of this community is the evaluation, not the profile. Members will review concrete artifacts — a preprint, an analysis pipeline, a proof, a study design — and score them on discipline-specific rubrics. Every rubric lives on this site, in public, before the first evaluation happens.

None of this replaces peer review of publications. It sits beside it, answering a narrower question journals were never designed to answer: can this person do rigorous research work, as judged by people qualified to tell.

How evaluation will work

The design centers calibration rather than authority. Evaluators will score artifacts independently; their scores will be compared against the pool and, where outcomes exist, against outcomes; and each evaluator's record of accuracy will accumulate on their own profile — per discipline, never globally.

That record is what will keep the system honest. An evaluator whose judgments consistently fail to hold up will carry less weight than one whose judgments prove out, whatever their title. Seniority can earn attention here; only calibration will earn weight.

Rubrics are public by design. A standard that cannot survive being read by the people it evaluates is not a standard worth scoring against. Each discipline page publishes the current rubric draft, and founding evaluators will pressure-test and revise it before it is applied to anyone.

All of this is written in the future tense deliberately. The mechanics — scoring, calibration records, evaluation panels — are designed and under construction, not live. Joining now means shaping them, not consuming them.

Why now

Research careers increasingly run through organizations that cannot read research. Deep-tech startups, applied labs, and funders all need to judge scientific ability, and most fall back on proxies — pedigree, publication venue, whoever will take a reference call.

A community that can evaluate research skill credibly, with published standards and accountable evaluators, is worth building carefully rather than quickly. This is the careful stage. The founding cohort exists to get the standards right before any volume arrives.

Eight disciplines at the start

Each page publishes the current rubric draft and what evidence of research skill looks like in that field. New fields will follow the standard, not precede it.

Questions worth asking first

Is the evaluation platform live today?
Not yet. The rubrics are public and the founding cohort is being assembled, but no evaluation has been scored on the platform. Joining now means joining the design stage: the first thing founding evaluators will do is stress-test the standards in their discipline, not work through a queue.
How much time does this ask of me?
Evaluation is episodic, closer in effort to refereeing a short paper than to sitting on a standing committee. During the founding stage the main work is critiquing the rubric drafts in your field, which fits in the margins of a normal working week.
Do I need a PhD or a faculty position to apply?
No. The application asks for evidence of research work we can read — publications, preprints, a substantial technical artifact, an ORCID or repository. Industrial researchers and early-career scientists are exactly the people the credential system undervalues, and that gap is part of why this community is being built.
Is founding evaluation paid?
No. What founding evaluators get is concrete but different: a real hand in setting the standards, a calibration record that starts on day one, and standing in a community built to make good judgment visible. If evaluation later carries compensation, that design will be worked out with the founding cohort, in the open.
What happens to the material I submit with my application?
It is used to review the application and for nothing else. The form links to public artifacts — ORCID, repositories, preprints — rather than asking you to upload unpublished work, and application material is not shared outside the review.

Apply to the founding cohort

The application asks for your discipline, an email address, and a link to work we can read — an ORCID profile, a repository, a preprint. Every application is read by a person; the bar is evidence of rigorous work, not affiliation.

VocaidDeep

If you hire researchers

Deep-tech teams that need to judge scientific ability have a different page — including how calibrated evaluation panels are being designed for hiring decisions.

Read the partner page

In design

Prizes funded by partners — and by the community itself

Evaluation tells you who did good work. It does not, on its own, give anyone a reason to show up. The missing half is funding: time-boxed problems worth solving, with a prize attached and a name on it.

  • Sponsored challenges

    A partner puts up a prize for a specific problem in health, AI/ML or research, and their name sits on it. Entries are judged against the same published rubrics as everything else on the platform — the sponsor funds the question, not the answer.

  • Virtual hackathons

    Scheduled, remote, and short. A weekend on one problem, judged openly, with the work left visible afterwards so it counts toward your track record rather than disappearing when the event ends.

  • Research programmes

    Longer-form work with a named partner and a named budget — for questions that will not fit in a weekend. Fewer participants, deeper commitment, published outcomes.

  • Community-funded pools

    Many small contributions gathered behind a problem the community picked itself, rather than one a sponsor chose. The community decides what is worth solving and funds it directly.

Four tracks, one rule

Challenges, hackathons and grant programmes run in four tracks. They differ in what the question is about — not in what they award. Whoever puts up the funds sets the question, never the score, and entering is free in every one of them.

  • Research

    Open questions, reproductions and methods work, plus articles worth commissioning. Judged against published criteria fixed before any funding attaches.

  • Health

    Problems scored from verified wearable metrics rather than self-reported effort, and only from data a participant has explicitly connected. Nothing here is medical: no diagnosis, no treatment, no clinical claim.

  • Career

    Skills, interview and project work that produces something an employer can actually look at. The evidence is shown to a company as evidence — never as a score that decides, and a person still makes the call.

  • Reputation

    The cross-cutting one: the quality of your calibration and your reviews. It is what the other three tracks feed, and it is earned by assessment rather than by paying to enter.

Vocaid never holds the funds

An external programme partner operates the funding and custodies it end to end — the prize goes from that partner to the person who won it. Vocaid runs discovery, participation and the record of what was done; it is not a payment intermediary and never takes custody of prize money. Entry is free in every track, so nothing you could lose was ever put in.

A track record you earn by review, not by reach

Articles, reproductions and challenge entries are scored by evaluators whose own calibration is on record. Ranking reflects work that peers have assessed — not follower counts, not posting volume, and not who paid to enter.

Paid in whatever is local to you

Funding a prize and receiving one should both work in the instrument you already use. Pix, card and stablecoin (USDC/EURC) are what Vocaid Pay names today; SEPA Instant, Wero and UnionPay are on the roadmap, not contracted. Prices and prizes are stated as amounts — Vocaid Pay is a way to pay, not a way to trade.

None of this is running yet. There are no live challenges, no prize pools and no participant numbers to quote, and there will be none until the funding, judging and payout rules are built and reviewed. Nothing here involves a token, and nothing you hold on Vocaid appreciates. If you would want to sponsor a problem — or bring a programme — the conversation below is the way in.

Join the VocAid community

Evaluators, hiring partners and candidates share one place — a Discord server with a space for each. Conversation, research previews and office hours happen there first, and it is where challenges and programmes will be proposed and argued over before any of them is funded.

Join us on Discord