A scoring system for invention disclosures, and the evidence under it.
Phaneris reads the public record around a university invention disclosure, scores it against earlier inventions with known results, and shows how every number was reached.
A method, a pipeline, and a record of results.
Every feature ranked among its peers
Each feature is read from a public record as it stood on the disclosure date and ranked among earlier inventions that have it, with the closest ones counting most. Intrinsic metrics read the disclosure's own record; comparative ones read the populations around it.
Eight public source families
Patents, company filings, university listings, publications, trials, court decisions, research funding and approved products. One query each, every URL recorded with the time it was read, and every node traceable back to the page it came from.
Checked against results
Scores that cannot be checked are opinions. As disclosures resolve, their results are read from dated events in public registers, and the scores are measured against them.
Offices that receive more disclosures than they can examine.
A research university receives several hundred invention disclosures a year, patents a fraction of them, and decides with the time and the evidence it has. The decision is usually made by a small team reading a form, under a filing deadline, with no comparable case in front of them.
The whole backlog, one standard
Every disclosure scored on the same evidence, so the twenty that already had a champion are not the only twenty examined.
Something to argue with
A number nobody can inspect is worse than no number. Every reading carries its rule, its source and its date, so a committee can disagree with a specific line.
One idea, one assessment
The same report for a single technology, without a licence or an account beyond an email address.
The limits, stated first.
No dollar figure, no royalty forecast. The scores rank a disclosure against earlier ones, nothing more.
A freedom-to-operate reading is a flag over public filings, not an opinion, and it says so wherever it appears.
A feature that cannot be found is left out and the rest averaged. Absence is reported, not filled in.
Raw disclosures stay in the institution's own instance. Only anonymised distributions leave it, and only once an office agrees.
Phaneris, from the Greek for visible, evident.
The product exists because the evidence for a decision is already public and scattered. Pulling it into one place, in the same shape every time, is the whole idea.