AlcorNotifire - AI-Powered Monitoring & Notification Platform

A modular AI-powered monitoring platform that continuously watches authorized data sources, analyzes new content using LLMs, and delivers intelligent real-time notifications through multiple channels.

Overview

AlcorNotifire is an AI-powered monitoring and notification platform designed to continuously monitor authorized online sources, analyze newly discovered content using Large Language Models (LLMs), and notify users when information matching their interests becomes available.

Unlike traditional keyword-based alert systems, AlcorNotifire combines scheduled content ingestion, AI-powered semantic analysis, event-driven processing, and multi-channel notifications to determine whether new information is genuinely relevant before notifying users.

The platform is built as a modular monolith, allowing clear separation of business capabilities while remaining simple to develop, deploy, and evolve. Each module owns its business logic and communicates through internal events, making the system highly maintainable and future-ready for service extraction if scaling requirements emerge.

Features

Modular Monolith Architecture

Feature-Based Modules: Authentication, Monitoring, Sources, AI Analysis, Alerts, Notifications and Background Jobs
Clear Module Boundaries: Independent business modules with well-defined responsibilities
Internal Event Bus: Decoupled communication between modules using domain events
Shared Infrastructure: Common configuration, logging, database access and utilities

AI-Powered Content Analysis

Semantic Relevance Detection: Goes beyond simple keyword matching
LLM Summarization: Generates concise summaries for detected content
Content Classification: Categorizes monitored documents
Configurable Prompt Templates: Easily customize AI behavior

Monitoring Engine

Scheduled Monitoring: Cron-based polling of external providers
Provider Architecture: Easily add new monitoring sources
Content Normalization: Standardizes data from different providers
Duplicate Detection: Prevents processing identical content multiple times

Notification System

Multi-Channel Delivery
- Email
- Slack
- Discord
Retry Mechanism
Delivery History
Channel Configuration

Background Processing

BullMQ Queues
Redis-backed Workers
Automatic Retries
Job Scheduling
Asynchronous AI Processing

Modern Dashboard

React + Vite
Monitor Rule Management
Alert History
Notification Configuration
Source Management
System Status Dashboard

System Design

Core Components

Monitoring Engine — Schedules monitoring jobs, manages monitoring rules, and coordinates external source providers.
Source Providers — Fetch and normalize data from RSS feeds and other authorized APIs using a pluggable provider architecture.
AI Analysis Engine — Evaluates content relevance, generates summaries, classifies documents, and produces structured analysis results using LLMs.
Alert Engine — Creates alerts from AI analysis results and manages their lifecycle.
Notification Engine — Delivers alerts through multiple notification channels with retry handling and delivery tracking.

Architecture Layers

┌───────────────────────────────────────────────┐
│         React Dashboard (Vite)                │
├───────────────────────────────────────────────┤
│             NestJS REST API                   │
├───────────────────────────────────────────────┤
│ Authentication │ Monitoring │ Alerts │ Users  │
├───────────────────────────────────────────────┤
│ Sources │ AI Analysis │ Notifications │ Jobs  │
├───────────────────────────────────────────────┤
│ PostgreSQL │ Redis │ OpenAI │ External APIs   │
└───────────────────────────────────────────────┘

Event-Driven Processing Pipeline

Scheduler

      │

      ▼

Source Provider

      │

      ▼

Normalize Content

      │

      ▼

Store Document

      │

      ▼

AI Analysis

      │

      ▼

Create Alert

      │

      ▼

Notification Queue

      │

      ▼

Email / Slack / Discord

Module Architecture

Module	Responsibility
Authentication	User authentication and authorization
Monitoring	Monitoring rules, scheduling and orchestration
Sources	External provider integrations
AI Analysis	LLM-powered relevance analysis and summarization
Alerts	Alert creation and lifecycle management
Notifications	Multi-channel notification delivery
Jobs	Background processing and queue management
Shared	Common utilities, configuration and logging

Technology Stack

Frontend: React 19, Vite, TanStack Query, Tailwind CSS, shadcn/ui
Backend: NestJS, TypeScript
Database: PostgreSQL, Prisma ORM
Queue: BullMQ with Redis
AI: OpenAI API
Infrastructure: Docker, Docker Compose, AWS (ECS, RDS, ElastiCache, S3, CloudFront)
Tooling: pnpm Workspaces, Vitest, ESLint, Prettier
Architecture: Modular Monolith, Event-Driven Design, Provider Pattern

Design Philosophy

AlcorNotifire is built around the principle that relevance is more valuable than volume. Rather than forwarding every piece of matching content, the platform uses AI to understand context, evaluate significance, and notify users only when information is genuinely relevant.

The architecture emphasizes modularity without operational complexity. A modular monolith provides clear ownership boundaries, independent feature development, and maintainable code while avoiding the deployment and operational overhead of microservices.

Business capabilities communicate through internal events rather than direct dependencies, allowing each module to evolve independently. External systems—including AI providers, content sources, and notification channels—are abstracted behind provider interfaces, making the platform easily extensible without impacting the core domain.

Designed as a production-ready backend platform, AlcorNotifire demonstrates modern software architecture practices including event-driven processing, asynchronous job execution, AI integration, modular design, cloud-native deployment, and clean separation of concerns.