The reference rate, RFP language, and audit trail your team needs to defend AI labor purchases internally.
A transaction-anchored index, an open methodology, and drop-in procurement artifacts. Buy with the same rigor you apply to commodity sourcing — without inventing the framework yourself.
Procurement runs on defensible references. AI labor doesn’t have one yet — so we built it.
For commodity categories, your team already has Bloomberg, Kelley Blue Book, SOFR, or an analyst feed your CFO trusts. For AI labor, there’s nothing. Vendor list prices are self-reported, analyst surveys are quarterly aggregates, and your in-house spreadsheet only reflects what you’ve already paid. None of that survives a serious sourcing review.
The WorkForce Labor Index is the missing artifact. Transaction-anchored medians for commodifiable AI work, computed with a published pipeline, IOSCO PD415-aligned, CC-BY-4.0 licensed, and republished as inputs clear an explicit input-reality gate. When a category isn’t ready, the median is held — not estimated. Your audit team will recognize the posture.
Around the index we publish the artifacts procurement teams actually need to ship work: RFP language with WLI-anchored pricing clauses, a reproducible vendor evaluation rubric, an end-to-end buying playbook, and a FinOps integration pattern for benchmarking AI spend against the index. You can copy these into your stack tomorrow.
What this is not: a managed procurement platform, an analyst subscription, or a vendor directory dressed up as a benchmark. It is the methodology, the rate, and the artifacts — open, cited, and reviewable.
Four drop-in artifacts for AI labor sourcing.
Each one is published under CC-BY-4.0, versioned in the public repository, and references the same underlying methodology.
RFP template
Drop-in RFP language for AI labor scopes — WLI-anchored pricing clauses, methodology disclosure requirements, audit terms, and termination triggers tied to gate failures.
Vendor evaluation framework
A reproducible scoring rubric for AI labor vendors. Versioned, open-licensed, designed for repeat reviews. Capability evidence, pricing transparency, evaluation outcomes — all weighted, all auditable.
Procurement playbook
End-to-end buying process for AI labor: intake, market sizing against WLI, RFP issuance, evaluation, pricing anchor, contract clauses, ongoing benchmark review. Steps, owners, artifacts.
FinOps integration
Pattern for benchmarking AI spend against published WLI medians. Variance reporting, anchor-rate clauses for renewal review, board-ready quarterly attribution against the index.
Why WLI is the reference rate procurement can defend.
Four properties your audit and legal teams will check for. Each is structural — not a marketing claim.
Transaction-anchored
Inputs come from verified transactions, evaluation outcomes, and structured vendor disclosures. No analyst panels, no buyer surveys.
Bootstrap-resampled
10,000 iterations per published median. Confidence intervals are published alongside the rate, not buried in a footnote.
Input-reality gated
Where a category fails the gate, the median is held — not estimated, not interpolated. Publication is the contract; gating is how it’s kept.
Open and cited
Pipeline published. License CC-BY-4.0. Public repository with version tags and audit log. Citable in board decks and regulatory filings.
What the institution stands on — evidence, not customer logos.
Honest posture. No claims out of scope.
Where a control is in scope but not yet attested, we say so. Where data residency is configurable, we say where.
The index is not a survey. It is a transaction-anchored, bootstrap-resampled reference rate.
Procurement through Workforce, against the alternatives.
The same five questions applied to each source likely already in your stack.
Procurement through Workforce (WLI-anchored)
Direct vendor contracts
Contractors and BPOs
In-house spreadsheet
Six surfaces. One methodology.
Each surface either feeds the index, depends on the index, or distributes it. No siloed feeds.
Marketplace
Where AI labor transacts — the Tier A volume that anchors the index as input-reality gates clear.
/marketplace →WLI
Published medians for cleared categories. Full audit trail. Citable in board decks.
/wli →Methodology
Anchoring, resampling, gating, governance. Open and versioned in the public repository.
/methodology →Compare
Side-by-side vendor and category comparison, anchored to evaluation outcomes and disclosed pricing.
/compare →Eval
The evaluation harness whose outcomes feed the inputs behind category medians.
/eval →Calculator
Model your unit economics against the index. Cleared categories only — no held medians.
/calculator/ai-agent-cost →Tell us what you’re sourcing.
Walk us through what’s on your desk — RFP, vendor consolidation, FinOps review, compliance gate. We route based on what you tell us. Real human reply, not a sequence.
Procurement-grade AI labor benchmarking — open, cited, and reviewable.
Open the form, tell us what you’re sourcing, and we’ll route a real reply within 24 hours.
Talk to procurement