All projects
Industrial SaaS · Multi-tenant Web Platform·2025

AssetGuard — Industrial Asset Monitoring SaaS

A multi-tenant industrial asset monitoring platform designed to track facilities, assets, and monitoring points in real time.

Open live project
ReactNext.jsTypeScriptTailwind CSSTanStack QueryNode.jsNestJSPostgreSQLWebSocketsZodJWT AuthenticationGitHub ActionsVercel

Overview

AssetGuard is an industrial monitoring SaaS built to simulate a real-world enterprise IIoT environment. The platform combines telemetry visualization, alert operations, health scoring, and maintenance insights so operations teams can monitor performance continuously and anticipate failures earlier.

Context

This project was built to simulate a real-world industrial monitoring platform inspired by enterprise IIoT systems. The focus was to validate a scalable architecture for multi-company environments, real-time data streams, and deep hierarchical relationships between companies, facilities, assets, and monitoring points.

Problem

Design and implement a production-style monitoring product that supports tenant isolation, role-based operations, real-time telemetry, and context-aware dashboards without sacrificing navigability as the industrial hierarchy grows.

Responsibilities

  • Designed and implemented multi-tenant architecture with role-based access for Platform Admin and Company Admin profiles
  • Structured hierarchical navigation across Companies, Facilities, Assets, and Monitoring Points with contextual routing
  • Built real-time telemetry simulation and WebSocket update flows for live operational visibility
  • Implemented an alert system with severity levels, status transitions, and filtering by operational context
  • Developed explainable asset health scoring rules with maintenance recommendation outputs
  • Created context-aware dashboards for global and tenant-specific monitoring perspectives
  • Implemented advanced filtering and search patterns to accelerate operational navigation
  • Maintained typed data contracts and validation boundaries with Zod in frontend and backend flows

Technical decisions

Tenant-aware domain model from the start

Tenant isolation and role boundaries were encoded at the domain layer, reducing cross-tenant leakage risk and keeping authorization rules explicit as modules evolved.

WebSocket-first telemetry channel

Telemetry streams were modeled as real-time event channels over WebSockets, enabling responsive dashboards and alert lifecycle updates without polling overhead.

Shared contracts between UI and API

A shared domain model with typed schemas kept frontend and backend aligned, reducing integration regressions across hierarchical resource views.

Architecture

Modular frontend and backend separation with shared domain modeling, real-time communication over WebSockets, and tenant isolation enforced through role-based access control.

Challenges

  • Balancing navigation clarity while traversing deep industrial hierarchies across multiple tenants
  • Keeping telemetry updates responsive under frequent event bursts while preserving dashboard readability
  • Designing health scoring and recommendation logic that remains explainable for operations teams

Results

Delivered a production-style SaaS workflow covering multi-tenant operations, live monitoring, and maintenance-oriented decision support

Validated a scalable architecture that cleanly separates tenant, domain, and real-time concerns

Established a deployable pipeline with GitHub Actions and Vercel for continuous iteration

Learnings

  • Industrial SaaS products benefit from domain modeling discipline early, especially for tenant and hierarchy boundaries
  • Real-time systems need explainable UX patterns so operations users trust automated scores and alerts
Back to projects