Ai chatbot

Title: Implement a secure AI chatbot for Venba ICU Registry (PICU, AICU and NICU) Venba Healthcare requires a developer to implement a secure AI Assistant within its ICU Registry web and Flutter applications. The product uses TypeScript/TSX, Flutter and a microservice architecture. Each hospital has a separate database and enables PICU, AICU, NICU or a combination during onboarding. The MVP will be piloted with one hospital but must be tenant-aware and reusable for future hospitals. The chatbot must use the authenticated user's hospital and existing RBAC/ABAC context, call only approved read-only APIs/data tools and answer a signed-off catalogue of 10-15 questions with accuracy as the highest priority. It must never execute unrestricted model-generated SQL or expose credential, user-security, configuration or another hospital's data. The assignment includes backend integration, TSX and Flutter UI, one AI-provider integration, testing, UAT support, documentation, deployment package and handover. Delivery will follow the five fixed-price milestones and acceptance gates defined in this specification