What we're building

Completed intake chatbot conversational flows

The AI intake chatbot now handles complete conversations for both Family Law and Estate Planning matters. We've tuned the conversation flows based on typical intake patterns and the specific information practitioners need.

• Practice area detection and appropriate question routing
• Document upload during conversation
• Conflict information gathering
• Structured summary output

Document generation MVP complete

First drafts of core documents can now be generated from intake data.

• Family Law: Consent orders (parenting), Consent orders (property), Parenting plans
• Estate Planning: Simple wills (NSW, VIC), Enduring power of attorney (NSW, VIC)

Clio integration working

The Clio integration now successfully syncs contacts and matters from Jurisdox. When you convert an intake to a client, the information flows through to Clio automatically.

Dashboard interface redesign

Redesigned the attorney dashboard for faster intake review. The new layout shows key information at a glance and reduces clicks to common actions.

• Intake queue with priority indicators
• Expandable summary cards
• Quick action buttons (convert, decline, schedule)
• Improved mobile layout

AI model improvements

Improved the underlying AI model configuration for better Australian legal context. The chatbot now handles Australian terminology more naturally and asks more relevant follow-up questions.

Conflict checking system

Implemented automated conflict checking based on names and entities collected during intake. The system flags potential conflicts before you engage with new clients.

Document generation architecture

Finalised the technical architecture for AI document generation.

• Using RAG (retrieval-augmented generation) to incorporate intake data
• Human-in-the-loop review required for all outputs
• Complete audit trail of all generations and edits
• Export to DOCX and PDF

Technical architecture finalised

After evaluating options, we've settled on our technical stack.

• Next.js frontend deployed on Vercel
• Supabase for database, auth, and storage
• Claude API for AI capabilities
• Australian region hosting for data sovereignty

Customer discovery complete

Completed customer discovery interviews with 18 Australian solo practitioners.

• Intake and document drafting are the most requested automation targets
• Price sensitivity is real—practitioners want tools under $100/month
• Contract lock-in is a major concern
• Integration with existing tools is preferred over replacement
• Trust in AI is cautious but growing—human review is non-negotiable