Skip to content

IoT

Connect Clean

Built an IoT Monitoring Platform from Scratch

Role
Full-stack developer
Timeline
Ongoing
Team
Small team
Live link
dashboard.connect-clean.com/dashboard
Connect Clean interface
24/7
continuous device ingestion
Multi-site
estate-wide monitoring

The problem

Businesses consume water and chemicals continuously across dozens of sites, but almost none of it is measured. Connect Clean needed smart dispensers that measure every use, a backend to ingest readings continuously and securely, and a dashboard where a manager anywhere could see every location and device at once.

My approach

Device connectivity & ingestion

A secure backend interface field-deployed dispensers report into continuously rather than on a polling schedule, handling devices dropping off and reconnecting so the record stays consistent.

Backend & data model

Node.js API and processing layer modeling organizations, sites, devices, cartridges, and time-series usage readings that roll up from device to site to estate.

Alerting

Real-time alerting on low supply, equipment faults, and anomalous consumption, with thresholds tuned to stay actionable rather than routine noise.

Reporting & analytics

Per-site and per-period consumption views, cross-location comparisons, and trends businesses use to identify waste and evidence sustainability targets.

Dashboard frontend

Next.js monitoring interface for facility managers: a single view across all locations, live status, alert handling, and report generation.

System shape

  1. Frontend

    Next.js, React, TypeScript

  2. Backend

    Node.js

  3. Data

    PostgreSQL

  4. Infra

    AWS

Screens

Connect Clean screenshot 2
Connect Clean screenshot 3
Connect Clean screenshot 4
Connect Clean screenshot 5

Stack

Frontend

  • Next.js
  • React
  • TypeScript

Backend

  • Node.js

Data

  • PostgreSQL

Infra

  • AWS

Result

A live production IoT platform connecting physical dispensers to a cloud dashboard, with continuous telemetry, real-time alerting, and usage analytics, running 24/7 against hardware deployed across multiple locations.

Next projectFastyr AI