The difference between holding files and holding a licensable training asset: rights, provenance, privacy, quality, packaging, and proof.

Holding a hard drive of logs is not the same as holding a training asset a frontier lab can license. “Lab-ready” means a buyer can answer yes to rights, privacy, provenance, quality, packaging, and proof—usually in that order. Highford publishes this checklist so sellers and buyers share a vocabulary before an introduction.

1. Rights that name AI training

Commercial use is not enough. Buyers ask whether the license or underlying contracts expressly allow model training, fine-tuning, and distillation, and whether redistribution of derived models is restricted (COMPEL). Silence in a pre-generative-AI contract is treated as risk, not consent. Copyright litigation has made the cost of ambiguity concrete: a federal court approved a $1.5 billion settlement in Bartz v. Anthropic over books used to train Claude, with per-work estimates reported around $3,000 before deductions (AP News).

2. Provenance you can defend

Labs want a chain from source system to delivered file: collection method, time window, transformations, exclusions, and who authorized the transfer. If a seller cannot map source-to-asset, serious buyers walk (Dayda; COMPEL). Provenance also feeds EU AI Act training-content summary expectations for general-purpose model providers.

3. Privacy that survives diligence

Unanonymized personal data is typically barred from marketplace listing (Datarade publishing rules). Health, biometric, and children’s data trigger sector rules. In California, businesses that knowingly collect and sell consumers’ personal information without a direct relationship may need Delete Act registration and DROP participation (CPPA). Lab-ready packaging usually means documented redaction, retention limits, and a sample that contains no residual identifiers.

4. Quality a grader can score

Static datasets need documented schemas, duplicate rates, class balance, and—if labelled—inter-annotator agreement. For reinforcement learning and agent training, quality increasingly means verifiable rewards: unit tests, database state checks, exact-match graders, or sanitizer crashes rather than only a preference model. OpenAI’s reinforcement fine-tuning docs expose string-check, Python, and score-model graders (OpenAI graders; RFT guide). Allen AI’s Tülu 3 formalizes reinforcement learning with verifiable rewards (RLVR) using deterministic binary verifiers (arXiv:2411.15124). DeepSeek-R1 emphasizes rule-based accuracy and format rewards for reasoning to reduce reward hacking (Nature). Anthropic’s engineering guidance on agent evals prefers code graders where outcomes are objective and full transcripts for replay (Demystifying evals for AI agents).

5. Packaging a pilot can consume

Lab-ready delivery is boring on purpose: stable schemas, checksums, version pins, sample under an evaluation license, acceptance criteria written before the pilot, and refresh SLAs for ongoing supply (SourceX). Boutiques that productize this work quote fixed fees from the low thousands for audits to mid-five and six figures for full data-product builds (licens.io). Highford does not perform that delivery; it helps counterparties discover who does—and who already holds licensable inventory.

6. Proof that survives a second run

Reproducibility is part of readiness. Buyers keep run records: model IDs, prompts, tool outputs, grader versions. Anthropic recommends multiple trials because stochastic systems move between runs, and regression suites that should stay near 100% pass (Demystifying evals). A lab-ready seller anticipates that culture: frozen dataset versions, published known issues, and the ability to re-ship the same bytes.

A short readiness scorecard

Gate	Fail signal
Rights	No written AI-training grant
Provenance	Cannot name source systems
Privacy	PII in the sample
Quality	No documented metrics or verifiers
Packaging	No evaluation license / acceptance tests
Proof	Cannot re-deliver the same version

Sellers who clear these gates earn faster diligence. Buyers who publish their gates get better inbound. Highford’s role is matching—searchable listings, AI matching, double-opt-in email intros, and agent APIs—not transferring the data itself.

What Highford will not do

Highford will not clean datasets, sign licenses on a party’s behalf, or take custody of files. The product is information and introductions: public-source listings remain free; people- or agent-submitted listings unlock with credits; matching and double-opt-in email intros connect parties who then contract directly. That boundary keeps the service on the Apollo-for-private-assets side of the line rather than the managed-delivery side occupied by labelling vendors.

Common failure modes sellers underestimate

Three failure modes show up often in public buyer guidance. First, rights nested in customer contracts: a SaaS company may own product telemetry in practice but have enterprise MSAs that forbid training use or onward licensing. Second, silent licenses: open-source or purchased corpora whose terms predate generative AI leave buyers exposed. Third, sample theater: a polished 1% slice that does not match the messy remainder fails the first pilot. Dayda’s guidance to insist on stratified samples—and SourceX’s insistence that evaluation licenses permit actual metric testing, not inspection-only views—exists because of those failure modes.

A fourth failure mode is verifier theater for interactive data: an environment that looks realistic but scores with a soft preference model the lab cannot reproduce. Buyers who have adopted RFT-style graders will ask for the Python check, the unit test, or the sanitizer build.

This page, for an agent

The same page as markdown, plus the API and MCP calls that belong to it.

API

Search public teasers

GET /api/v1/listings?q=unpaid+judgments+in+New+York

Create a free account

POST /api/v1/signup
Content-Type: application/json

{"email":"ada@example.com"}

api_key and token come back immediately. Search works with that key. Unlocks, intros, and credits wait until the sign-in link is opened. POST /api/v1/verify with {"email":"ada@example.com","code":"<6 digits>"} still confirms the address.
A new account includes unlimited search, 50 contact unlocks a month, and 5 intros a month once the email is confirmed.

MCP

MCP server

{
  "mcpServers": {
    "highford": {
      "url": "/api/mcp"
    }
  }
}

Search

search_deals {"q":"unpaid judgments in New York"}

JSON

{
  "site": "Highford",
  "page": "/research/what-makes-data-lab-ready",
  "markdown": "/research/what-makes-data-lab-ready?format=md",
  "api": [
    {
      "label": "Search public teasers",
      "request": "GET /api/v1/listings?q=unpaid+judgments+in+New+York"
    },
    {
      "label": "Create a free account",
      "request": "POST /api/v1/signup\nContent-Type: application/json\n\n{\"email\":\"ada@example.com\"}\n\napi_key and token come back immediately. Search works with that key. Unlocks, intros, and credits wait until the sign-in link is opened. POST /api/v1/verify with {\"email\":\"ada@example.com\",\"code\":\"<6 digits>\"} still confirms the address.\nA new account includes unlimited search, 50 contact unlocks a month, and 5 intros a month once the email is confirmed."
    }
  ],
  "mcp": {
    "url": "/api/mcp",
    "calls": [
      {
        "label": "MCP server",
        "request": "{\n  \"mcpServers\": {\n    \"highford\": {\n      \"url\": \"/api/mcp\"\n    }\n  }\n}"
      },
      {
        "label": "Search",
        "request": "search_deals {\"q\":\"unpaid judgments in New York\"}"
      }
    ]
  }
}

Markdown

Reading this page…

What makes data lab-ready