The regulatory copilot
Ask the law your question. It answers, article in hand.
ERP classification, occupancy and category, accessibility, fire safety: Lex Chat answers in plain language, citing the exact texts. Not a generic chatbot: an expert engine for French building law.
A 180-seat restaurant with takeaway sales: which ERP classification, which category?
Multi-activity establishment: type N (dining) + M (retail), main activity N. Public occupancy retained: 180 people → 4th category.
And does staff headcount matter?▍
You ask
A trade question, in plain words
“Which classification for a restaurant with takeaway sales?”, the way you'd ask a fire-safety specialist.
The engine rules
The encoded rule executes
The question is routed to the competent expert engine: classification, occupancy, accessibility or fire. The rule is applied, not guessed.
You decide
A sourced, defensible answer
Every conclusion cites its articles: CCH, orders, safety regulation. Enough to decide, and to prove it.
What it changes
Automatic ERP classification
All 14 establishment types, including multi-activity venues (main activity, secondary activities, composite type). The engine walks the regulatory decision tree.
Occupancy & category
Maximum occupancy (public + staff) and category 1 to 5 per article R. 143-19, activity by activity, level by level.
Accessibility compliance
Paths, ramps, lifts, restrooms, adapted parking: the four reference orders covered, for existing buildings and new construction alike.
Sourced fire-safety answers
Fire resistance, smoke extraction, egress: a research agent searches the raw regulatory corpus and cites every article it uses.
Zero hallucination, by design
Expert-engine answers are generated from deterministic templates: the system can only cite what is actually in the rule.
A conversation that follows the project
Lex Chat remembers the building under study and follows up: “and this ERP's occupancy?” reuses the previous classification, like a colleague would.
The SynopticLex suite