How mukandkrishna/eyecon-ai works
Project Type and Purpose
• This is a full-stack, microservice-augmented anomaly-detection platform, branded “Eyecon AI.”
• It consists of:
• A Next.js/TypeScript frontend for admin, super-admin, public demo, and auth flows.
• An Express-based Node.js backend exposing REST endpoints for anomalies, cameras, organizations, users, analytics, etc.
• A Kafka cluster + Python producer for ingesting streaming data (e.g., video or telemetry) into the system.
• A relational database (via config/database.js) and Firebase services (config/firebase.js) for auth/storage.
A. Frontend (Next.js App)
• Runs in Vercel or Docker.
• Uses NextAuth, React contexts, TailwindCSS, custom UI library, and calls backend API routes under /api.
B. Backend API (Node.js + Express)
• Dockerized, composed via docker-compose.yml.
• Layers:
• Routes (e.g. routes/adminRoutes.js, publicRoutes.js)
• Controllers (e.g. anomalyController.js)
• Services (e.g. GroqService.js)
• Models/ORM (e.g. Sequelize models under models/)
• Middlewares (auth, org-access)
• Connects to: relational DB, Firebase, and optionally a Kafka consumer.
C. Database
• A SQL database defined by config/database.js (Postgres/MySQL).
• Seeded via scripts under scripts/.
D. Firebase
• Used for auth, storage, real-time features via config/firebase.js.
E. Kafka Messaging
• Kafka-Server (broker + zookeeper) launched via its own docker-compose.yml.
• Kafka-Producer (Python) pushes messages into the Kafka topic(s).
• (Optional) A backend consumer microservice reads from Kafka and writes to DB or triggers anomaly analyses.
F. External/3rd-party Services
• Groq (or other AI/analytics engines) via GroqService.js.
• Firebase Auth/Storage.
Relationships and Data Flows
• Backend → Firebase: SDK/API calls.
• Producer → Kafka: Python producer publishes messages to Kafka topic(s).
• Backend → Groq/External AI: HTTP/gRPC calls for analytics.
• Frontend ←→ Firebase: (if real-time or storage is used directly by frontend).
Architectural Patterns & Design Principles
• Layered MVC in Backend: clear separation of routes, controllers, services, models, utils.
• “Monorepo” structure with distinct subfolders for backend, frontend, and Kafka services.
• Microservices/Message-driven ingestion (Kafka).
• Containerization: each major piece runs in its own Docker container, orchestrated via Docker Compose.
• Next.js App Router with API routes and server/client components.
Essential Elements Checklist
[ ] Clearly labeled components (e.g., “Next.js Frontend,” “Express API Server,” “Postgres DB,” “Kafka Broker”)
[ ] Callouts for key tech: Node.js, Express, Sequelize, Next.js, Kafka, Docker, Python, Firebase