ResQConnect Documentation

Welcome to the ResQConnect technical documentation. This site provides detailed information about the system architecture, data models, API endpoints and workflows.

Overview

ResQConnect is an end-to-end, agentic disaster-response decision support system that turns incoming citizen requests into actionable guidance and allocations of emergency resources. It combines an agentic RAG pipeline (to retrieve and synthesise verified SOPs and local knowledge into safe, context-aware recommendations), a dynamic routing/dispatch optimisation layer (to prioritise incidents and assign responders under changing constraints) and an edge/offline component (so core guidance remains available when connectivity is limited). The goal is to improve timeliness, safety and consistency of response decisions while keeping outputs under human supervision .

Sections

System Overview: End-to-end mental model of how frontend, backend, and mobile connect.
Codebase Map: Where major responsibilities live in each codebase.
Dataflow: How requests, workflows, tasks, and resources move through the system.
Agentic Workflow: AI planning lifecycle with moderation and execution states.
Chatbot Architecture: Technical flow for /chatbot/ask and context-aware responses.
Offline LLM: On-device inference path and mobile online/offline sync model.
Backend Architecture: High-level overview of the backend system.
Data Models: Entity-Relationship diagrams and schema definitions.
Sequence Diagrams: Visual representation of key system workflows.
API Reference: Detailed API endpoint documentation.
Frontend Overview: Introduction to the frontend application.
Frontend Architecture: Project structure and key patterns.
UI Components: Design system and component organization.
API Integration: Data fetching and service layer details.
Mobile Overview: Flutter app goals, stack, and current capabilities.
Mobile Architecture: Under-the-hood runtime and bridge internals.
Mobile Workflows: Model download, inference, and help-request flows.