AI policy
AI is limited to revenue, intake, and attribution intelligence.
Case Yield AI helps match messy inquiries to signed cases, explain evidence, classify intake quality, and recommend cut, scale, or fix moves.
Allowed AI use
- Source Match AI assigns confidence and evidence for messy inquiry-to-signed-case matching.
- Intake Quality AI flags urgency, disqualifiers, language, practice area, and follow-up quality.
- Ask Case Yield answers grounded analytics questions from the firm's own dashboard data.
Guardrails
- AI must not provide legal advice, legal strategy, or attorney-client communications.
- AI must not create an attorney-client relationship or decide whether a person has a legal claim.
- Sensitive decisions require human review and source evidence, not model output alone.
Transparency
- Recommendations include evidence, reason codes, confidence, and the data behind the finding.
- Attribution overrides and manual review decisions stay tied to source evidence and reviewer context.
- Model usage stays off until an operator-approved provider configuration is present for the deployment.
Data protection
- Provider approval is handled through customer and operator configuration before firm data is processed.
- Prompts are minimized, direct identifiers are redacted where practical, and untrusted lead facts are fenced as data.
- AI-generated memos are business analysis, not legal work product unless separately reviewed by counsel.
Case Yield supports revenue attribution and intake intelligence only. It does not provide legal advice, create an attorney-client relationship, or make legal decisions for a firm.