
Design an AI-first intake interview form
You'll have a wireframe or clickable prototype of an intake form where an AI voice agent (not a human) conducts the initial interview, captures structured data, and hands off only the gaps to your tea
What you'll make
AI Voice Intake Form Wireframe
- Opening Screen: voice consent and category risk disclosure
- Question Sequence: intake questions designed to expose each failure category
- Validation Checkpoints: specific triggers for fabricated citations, wrong jurisdiction, and contradictions
- Failure Handling: how the interface responds when each validation category fails
Upload screenshots or PDF export showing the complete voice interaction flow with validation logic
The 4 parts
Each part ends with a check you can see.
Map Legal AI Failure Exposures
Test intake questions against the three legal AI failure categories to find which questions expose which risks
Done when
A risk map showing which intake questions expose fabricated citations, wrong jurisdiction, or internal contradictions, with specific examples
Design Category-Specific Validation
Build validation checkpoints that catch each legal AI failure category with specific trigger conditions
Done when
Three distinct validation checkpoint designs with concrete trigger rules and escalation workflows for each failure category
Test Voice Interface Failure Handling
Create and test voice prompts for when each validation category fails, measuring user experience impact
Done when
Tested voice prompts for each failure category with user feedback and refinement based on actual failure scenarios
Build the Complete Voice Wireframe
Assemble all components into a clickable wireframe showing the complete AI voice intake flow with validation
Done when
A complete wireframe demonstrating the voice interaction flow, all validation checkpoints, and failure handling for each legal AI category
Capstone: you put together your AI Voice Intake Form Wireframe.
Where this helps
An elective: a single lesson that sits outside rtwo's job certificates and doesn't count toward one.
About this lesson
What you'll be able to do
- Build a clickable wireframe of an AI voice intake form with validation rules for each legal AI failure category
- Test your validation checkpoints against 3 real legal AI hallucination scenarios and refine the triggers
- Have documented voice interaction patterns that handle the specific failure modes of legal AI
- Know which intake questions expose fabricated citations, wrong jurisdiction, or internal contradictions
Who it's for
- Solo practitioners and small firm partners who handle initial client consultations
- Legal tech coordinators designing client-facing automation
- Anyone who's wondered if AI could pre-qualify leads without creating malpractice risk
What it won't cover
- Bar compliance review for your specific jurisdiction
- Backend voice AI technical implementation or hosting infrastructure
- Client confidentiality agreements or engagement letter generation
- Integration with existing practice management systems
Skills you'll practise
- Category-specific validation design for legal AI
- Voice interaction failure mode mapping
- Hallucination scenario testing
- Legal AI risk assessment