Contract data from transactional attorneys.
Contract models need to read like a careful lawyer, defined terms and all. We commission vetted transactional attorneys to label, redline and grade the data they learn from.
Contract law, with proof.
Contract AI makes characteristic mistakes. It extracts a limitation of liability clause but misses the carve-out two sections later, reads a capitalized term without checking its definition, assumes US law in an English-law agreement, or suggests a redline that quietly shifts risk to the wrong party. HUMXN commissions credential-checked transactional attorneys and in-house counsel to create clause labels, redlines, playbook reviews and evaluations that teach models to read agreements as a whole.
Contracts are sourced only from consenting sources, with consent bound to each item, or drafted new by attorneys for the purpose. Documents are scanned for personal information and masked, fingerprinted at intake and checked for duplicates, and each label records whether it came from the author, a model suggestion or an expert reviewer. A second attorney reviews every item, and deliveries carry a signed receipt and audit trail your legal and procurement teams can verify.
For attorneys, this is paid, remote work in the area you already practice. You label clauses, write redlines with explanations, grade model analysis and review a colleague's work, at a rate shown before you accept each job. The work creates data for AI systems. It is not legal advice to anyone and forms no attorney-client relationship. Your identity is not published, and approved work is paid automatically through Stripe.
Built to your specification.
Every engagement starts from a written spec and a pilot batch. These are the most common requests we source for contract law.
Clause extraction and classification
Clause-level labels for indemnities, limitations of liability, termination, assignment, governing law and more, with defined terms resolved.
Redlines with rationale
Attorney-written markups against a stated playbook or party position, each change explained in plain language.
Playbook review data
Agreements reviewed against a negotiation playbook, flagging deviations, risk level and the fallback position an attorney would take.
Cross-jurisdiction sets
Parallel items under different governing laws, so models learn where enforceability and drafting conventions diverge.
Model answer grading
Attorneys score model summaries, extractions and suggested edits for accuracy, completeness and missed risk.
Drafted agreements
Newly drafted contracts and clause libraries written for training, avoiding the confidentiality issues of real client documents.
01Defined terms and cross-references
The meaning of a clause often sits elsewhere in the agreement. Attorneys label in context, not sentence by sentence.
02Jurisdiction differences
Enforceability and conventions vary by governing law. Data tagged by jurisdiction keeps models from assuming one legal system.
03Confidential documents
Real contracts carry client secrets. Consent bound to each item, PII masking and drafted-for-purpose agreements keep sourcing clean.
04Subtle risk shifts
A one-word redline can move liability. Second-attorney review catches edits that read well but change the deal.
Paid work in your field.
Remote and flexible, with the rate stated before you accept. Every job is reviewed by a second expert, and approved work is paid automatically.
- A JD, LLB or equivalent law degree
- Admission to practice in at least one jurisdiction
- Several years of transactional or in-house contracts experience
- Experience with specific agreement types, such as SaaS, supply, licensing or M&A
- Senior paralegals and contract managers for selected extraction tasks
Label clauses
Identify and classify clauses, resolving defined terms and cross-references across the agreement.
Redline against a playbook
Mark up agreements from a stated position and explain each change in a sentence or two.
Grade model analysis
Score model summaries and suggested edits for accuracy, missed risk and jurisdictional errors.
Review a colleague's work
Act as second attorney on another expert's labels and redlines before delivery.
Contract law: asked often.
Which jurisdictions and agreement types do you cover?
Requests are scoped by governing law, agreement type and party position, then matched to attorneys admitted and experienced in that area. Common work includes commercial, SaaS, licensing, supply, employment and M&A agreements, but the scope is set by your specification. A pilot batch confirms the labeling guide and standard before the request grows.
Where do the contracts come from?
Either from consenting sources, with consent bound to each item, or drafted new by attorneys for training. We do not use scraped or client-confidential documents. Every document is scanned for personal information and masked where needed, fingerprinted at intake, checked for duplicates and signed with provenance, so you can trace each item to its source.
Does this work create an attorney-client relationship?
No. You are creating and reviewing data for AI systems. You give no legal advice to anyone, represent no party and form no attorney-client relationship. Agreements you review are training material, and your analysis is used only as labeled data. Each job states its purpose before you accept.
Do I need to be admitted to the bar?
Most contract jobs require a law degree and admission in at least one jurisdiction, which we verify along with your identity. Some extraction tasks are open to senior paralegals and contract managers. Each job lists the credentials and jurisdiction it requires, and you are only offered work that matches your verified profile.
How do I get paid, and how much time is required?
Rates are set by field and task and shown before you accept. There is no minimum commitment: you choose jobs that fit around your practice and work remotely within each deadline. Work approved on second-attorney review is credited to your balance and paid automatically through Stripe once your identity and tax details are verified.
Can we get exclusive rights to the dataset?
Yes. Commissioned contract datasets can be licensed exclusively, with scope, term and exclusivity agreed in writing before work begins. Deliveries come in JSONL, CSV, Parquet or Croissant with a signed receipt listing every item by ID and hash, and standing orders can extend coverage to new agreement types.
Contract law, done by experts.
Need expert-made data in this field, or have the expertise to make it? Start here.