tungsten.dev

Customer case study · AI rights

Find media rights with the agreement language attached.

Legal experts taught Tungsten to apply the company’s own rights judgments across a large contract library. The agent returns a determination with evidence and sends ambiguity back to the experts.

Operating baseline

200K
contracts in the target population
100
paralegals in the manual alternative
1 year
estimated manual review

What the experts teach

Legal experts teach the company’s rights judgments.

  • The contract library and approved source material.
  • Work-for-hire provisions and their effect on rights.
  • The rights the company wants to identify and monetize.
  • Renewal terms, date calculations, and reviewed edge cases.

What the agent does

Tungsten applies each judgment and escalates ambiguity.

  • Interpret agreement language in context.
  • Return a rights determination with supporting evidence.
  • Escalate ambiguity instead of guessing.
  • Learn from reviewed corrections.

Agent output

A determination the legal team can inspect.

determinationRights determinations structured for the rights-management system.
evidenceThe supporting agreement language remains attached.
exceptionAmbiguous cases return to the legal expert.

Customer outcomes

Know the rights sooner. Keep legal review where it matters.

01

Rights known sooner

Identify which owned media rights can be monetized.

02

Manual hours avoided

Replace the manual hours implied by a year-long, 100-paralegal review.

03

Corrections carry forward

Expert feedback teaches how work-for-hire and renewal terms change the answer.

Start with one judgment-heavy task

Teach the first rights decision.

Bring a legal expert, representative agreements, and the decisions the business needs to make.

tungsten.dev