A registry of human-built rubrics and a settlement layer for measurable improvement. Evaluate artifacts against transparent standards. Rewards verified progress rather than estimated effort.
Hourly billing existed because quality was hard to quantify. Xantus evaluates the document itself — where it started, where it ended, and measures what changed in between.
Every Xantus interaction passes through three layers. Each layer serves a single purpose and stays focused on that role.
Score an artifact against a certified rubric.
An artifact is uploaded. A rubric — a human-authored measurement instrument with defined grading criteria, weighted categories, and scoring anchors — produces a per-criterion score, a written justification, and a normalized result on a 0–100 scale. Every evaluation is recorded as a public ledger event.
Two parties bind to a rubric and a baseline.
A payor and a payee enter a Lift agreement: a starting score, a target or per-unit improvement formula, a face value, and a deadline. The rubric becomes the neutral standard. Neither party can alter it. Neither party owns it.
Money moves on the measured delta.
When the artifact is re-scored, the contract auto-settles against its terms. The payor pays for measured lift. The payee earns based on what changed. The rubric author receives a share from every official evaluation. Every settlement remains fully auditable.
Every rubric in the Catalog is human-authored, version-locked, and reviewable by anyone. The criteria are explicit. The weights are public. The scoring guidance is on the page.
Pick a rubric. Upload an artifact. Get back a structured evaluation with a per-criterion scorecard and a written justification for each score. The result lives in a public Vault — referenceable, durable, audit-grade.
Rubrics are scoped to a domain: writing, business plans, technical specifications, performance reviews, supplier assessments. Each version is immutable. When the author improves the rubric, it becomes v2 — never a silent edit of v1.
Every run pays the rubric's author. Authors keep the majority share. The platform takes a small cut to maintain the registry. The split is published with every result.
Xantus does not price value; it measures lift.
The artifact owner escrows the agreement amount when the contract is accepted. If the payee never delivers measured improvement, the escrow is refunded.
The worker does the improvement offline, using the published rubric to iterate. Payout fires when the re-scored artifact meets the agreement terms.
The rubric author receives their per-run fee on each evaluation and has no stake in the agreement outcome. They were paid once, when the rubric was published.
No timesheets. No “good faith effort.” No subjective debate over whether the work was worth its price. The measurement decides.
Start free. Move to full participation when you're ready to author, transact, or run private evaluations.
For browsing the registry and running public evaluations.
For full participation in the Xantus economy.
.png)