Open Lluna

LegalQ AI Legal Category & Consultant Finder

A chat-first assistant that understands your situation, classifies the legal matter, and recommends the right category and vetted sources to find a consultant or lawyer.

85%+
Question Deflection (target)
PII-Safe
Redaction & Logging
24/7
Availability
RAG
Cited Sources
Problem

Legal information is fragmented and hard to navigate

  • Unclear category: Users don’t know if their case is civil, labor, real estate, business formation, IP, family, or something else.
  • Discovery fatigue: Searching laws, blogs, and forums yields conflicting answers.
  • Mismatched experts: Contacting the wrong consultant wastes time and money.
  • Privacy anxiety: People hesitate to share sensitive details online.
Navigating legal questions is confusing — LegalQ streamlines the path
Our Approach

Classify first, then guide with sources and next steps

We co-designed with legal professionals. The assistant extracts entities from the conversation, classifies intent into a legal category, and proposes vetted directories & resources—always with guardrails.

Understanding the case

  • NER over the chat (parties, dates, locations, contracts)
  • Intent + category prediction (e.g., labor dispute, tenancy)
  • Follow-ups to resolve ambiguity

Routes & resources

  • Recommends the right consultant category
  • Suggests official portals & trusted directories
  • Explains paperwork & prerequisites

Safety & compliance

  • PII redaction and safe-completion policies
  • Citations and disclaimers (not legal advice)
  • Audit logs for moderation
Solution

AI copilot with cited answers and the right next step

Built for clarity: short, sourced summaries plus the quickest path to a relevant professional.

Domain-tuned reasoning

Prompts and classifiers tuned for legal taxonomy. The bot asks clarifying follow-ups before suggesting actions.

IntentNERClarifying questions

Cited, plain-English answers

Summaries reference source passages so users can verify. Disclaimers keep the tone helpful but clear: not legal advice.

SummariesCitationsDisclaimers

Next best step

Maps the case to the right consultant category and links to vetted directories and portals to contact a human when needed.

RoutingDirectoriesHandoffs
Tech Stack

LLM + RAG with strict guardrails

Practical choices for safety and adaptability across jurisdictions.

AI Core
  • LLM orchestration (prompt templates, tools)
  • Zero-/few-shot classifiers (legal taxonomy)
  • Follow-up question planner
Retrieval (RAG)
  • Vector DB (policies, FAQs, laws)*
  • Hybrid search (keyword + semantic)
  • Chunk citations & score thresholds
Product
  • Next.js front-end (chat UI)
  • Streaming responses & feedback
  • Contact handoff + directory links
Safety
  • PII redaction in logs
  • Abuse filters & rate limits
  • Audit trail / safe completions
Impact

Help people act faster with confidence

↓ time to action
Users know their category and what to do next within minutes.
↑ consultant fit
Better matches reduce back-and-forth and no-shows.
↑ trust
Citations and disclaimers make guidance clear and verifiable.

Bring LegalQ to your users

Classify → explain → connect. A better way to start any legal journey.

Case Study path: /case-studies/legalq