IoT & AI platform for every domain · web & mobile · API-first
See every asset. Trust every reading.
Pragma IoT Portal models your sites, assets and devices — from production lines and power plants to farms, rivers and city streets — ingests measurements from devices, APIs and files, watches device health and data quality around the clock, and turns time-series data into dashboards, scheduled reports and a clean REST API. Then it lets your team act: rules that notify the right people, a live map, a control panel that sends commands, and a mobile app in every pocket.
Seeing the data is half the job. Rules turn conditions into alerts for the right people at the right time, the map shows where things happen, and the control panel lets operators respond.
Rules
Any tag, as latest value or as an average, sum, minimum, maximum, count or change over a window — compared with a number or with another tag, combined with AND / OR. Limit the rule to certain days and hours, notify everyone or selected members with the severity you choose, once per episode or repeated, capped per day, with an all-clear when it recovers.
Pump pressure too highWarning · active
if average pressure of Pump station 3 over 15 minutes is at least 3 bar
and latest pump running of Pump station 3 is equal to 1
Mon–Fri 09:00–17:00once per episode · max 5/dayoperators · e-mail + push
Triggered · 3.4 bar ≥ 3 · 10:42
Cleared · 2.7 bar · 11:05
Map
Pin assets where they are and choose per pin what to show: latest readings or trailing aggregates of any device tag. Pins take the colour of device health, cluster as you zoom out, and follow the light or dark theme. Members see the maps administrators build.
Plant overview · 5 assets
Pump station 3
1 device offline
Pressure · avg 1 h
2.7 bar
Flow · sum today
688 m³
Control panel
Design a card once per device type: which values to show, and which controls — toggle, setpoint, status, action button. Each control sends a templated payload to your gateway over HTTP (with authentication) or MQTT. Filter by type, asset or group; every command is logged.
Pump station 3 controllerseen 12 s ago
Pressure
2.26 bar
Flow
41 m³/h
PumpOn
Setpointnow 2.5 bar
3.0Send
Mode
AutoManualMaintenance
via MQTT plant-a/pump-3/cmd · last command ✓ 2 min ago
Nexus mobile app
The whole platform, in every pocket
A native app for Android and iOS built on the same API. Sign in, switch tenants, and the app takes on your tenant's logo and colours.
Alerts with push
Rule and device alerts arrive as push notifications; each person chooses which types reach the phone.
Devices, health, quality
Devices grouped by assets and groups, with health status, coverage, gaps and anomalies per device.
Dashboards & Explore
Read-only dashboards and the full Explore: pick tags, ranges, methods; share CSV or an image.
Map & control panel
The same maps and control cards as the portal, so operators can look and act from the field.
Distributed by your organisation. Ask your tenant administrator for the app; sign in with the same account as the portal.
pragmaNexusAcme Industries ⇄
Hello Olivia
Acme Industries
Online
41 / 42
Unread alerts
2
Dashboards Explore Map Control panel
Latest alerts
Pump pressure too high · 3.4 bar ≥ 3 · 4 min ago
Boiler B-2 stopped communicating · 18 min ago
HomeDevicesDashboardsExploreAlertsMore
AI ready · modular
Built for your team today, and for AI agents tomorrow
The portal is not a monolith. Each capability is a module behind one typed, tenant-scoped API — which is exactly what makes it safe to hand to an AI agent.
Typed contracts agents can call
Every capability is an OpenAPI endpoint with zod schemas — assets, devices, measurements, aggregates, dashboards, reports. No screen-scraping, no bespoke integration.
Agents act like people
An agent signs in with a role and an access scope, so it sees exactly the assets and devices it is allowed to — and every action lands in the audit log.
Structured events to reason about
Device offline, abnormal value, data-quality drop, import finished: events carry the context an agent needs to investigate, explain or escalate.
Pragma Agentic Framework
Built to plug into Pragma's agents: ask questions in plain language, get anomalies explained, have reports drafted and rollups tuned — through the same API your team uses.
Platform modulesone API · one tenant boundary
Asset & device model
hierarchy, specs, groups
Ingestion
devices, connectors, imports
Time series
TimescaleDB, rollups, aggregates
Monitoring & quality
health, anomalies, KPIs
Rules
conditions, schedules, throttling
Dashboards & reports
widgets, PDF schedules
Maps
pins, live values, clusters
Control panel
templates, HTTP / MQTT commands
Notifications
in-app, e-mail, push
Roles & access
custom roles, asset scopes
Mobile app
Android & iOS, branding
AI agents
Pragma Agentic Framework
REST API · OpenAPI · events — every module above is reachable through the same contracts, for people, integrations and agents alike.
Capabilities
Everything between the sensor and the decision
From the moment a device sends its first reading to the report in a manager's inbox, one platform covers the whole chain.
AI ready
Every capability is a typed, tenant-scoped API with structured events, so AI agents can query data, explain anomalies and draft reports with exactly the access a person would have. Built to connect to the Pragma Agentic Framework.
Modular by design
Asset model, ingestion, time series, monitoring, dashboards, reports, notifications, access control and AI are independent modules behind one API. Use what you need today and switch on the rest later.
Asset & device model
Sites, fields, buildings, lines and machines in a hierarchy; devices typed by specifications that declare every tag, unit and limit. Group anything, any way.
Ingest from anywhere
Per-device keys for direct telemetry, scheduled HTTP connectors for third-party APIs, and CSV / Excel imports with automatic schema detection.
Dashboards your teams build
Line, area, bar, value, pie, donut, heatmap and table widgets on a drag-and-resize grid, with calculated tags from expressions and per-widget time ranges.
Scheduled PDF reports
Compose reports from headings, text, images and the same chart widgets. Download as PDF or e-mail them daily, weekly or monthly to the people who need them.
Device health monitoring
Learns each device's reporting rate and alerts when it goes silent, recovers, or sends values outside its limits. Fixed thresholds when you prefer.
Data-quality KPIs
Coverage, gaps and data loss, spikes, stuck sensors and counter resets — per device and per tag, with a quality score you can act on.
Fast at scale
TimescaleDB under the hood with per-device continuous aggregates you configure yourself, so year-long queries answer in milliseconds.
Notifications that respect people
Devices, rules, imports, connectors and reports raise notifications in the portal, by e-mail and as push on the phone. Every member chooses what reaches them, per type and per channel.
Rules you write in plain terms
“If the 15-minute average pressure is above 3 bar and the pump is running, on weekdays 09:00–17:00, warn the operators — once per episode, at most five times a day.” Any tag, any aggregate, AND / OR, compared with a number or another tag.
Your assets on a map
Pin sites and machines where they are; each pin shows the values an administrator chose — latest reading or a trailing average, sum, min or max — coloured by device health and clustered as you zoom.
Control panel
Card templates per device type turn tags into live values and controls: toggles, setpoints, status switches and action buttons that send commands to your gateway over HTTP or MQTT, with a full command log.
Nexus mobile app
Android and iOS: alerts with push, devices grouped by asset, health and data quality, dashboards, explore, map and control panel — themed with your tenant's logo and colours.
Custom roles & access scopes
Compose roles permission by permission, and restrict any member to specific assets and devices. Every list, chart and alert respects the scope.
Domains
One platform, every domain that runs on measurements
The asset hierarchy, device specifications and time-series engine are generic. Name your sites, assets and tags the way your domain does and the rest of the platform follows.
Industry
Production lines, machines and utilities in plants: OEE inputs, energy per unit, vibration and temperature trends, downtime detection.
Manufacturing
Process plants
Maintenance
Energy
Substations, generation units, meters and grid assets: load profiles, power quality, counter readings and consumption reports.
Grid
Metering
Power plants
Renewable energy
Wind, solar and hydro parks: per-turbine and per-string production, availability, irradiance and yield against forecasts.
Wind
Solar PV
Hydro
Environment
Weather stations, air and water quality probes, noise and level sensors: regulatory reporting with data-quality guarantees.
Air quality
Weather
Hydrology
Agriculture
Soil moisture, greenhouse climate, irrigation and livestock buildings: field-by-field hierarchies and season-long histories.
Irrigation
Greenhouses
Silos
Smart cities
Street lighting, traffic, parking, waste and public buildings: district-level views for operators and open data for citizens.
Lighting
Mobility
Public assets
Water & utilities
Pumping stations, reservoirs, networks and treatment plants: flows, pressures, levels and leak indicators across the network.
Water
Gas
District heating
Buildings & facilities
HVAC, metering and indoor climate across a portfolio of buildings: comfort, consumption and equipment health per floor and zone.
HVAC
Energy management
Campuses
Logistics & transport
Fleets, cold chains and warehouses: telemetry from vehicles and containers alongside fixed infrastructure.
Fleet
Cold chain
Warehouses
Maritime
Vessels, ports and offshore assets: engine and fuel telemetry, cargo conditions, quay and crane activity, sea state.
Vessels
Ports
Offshore
Food & beverages
Production lines, cold rooms, fermentation and packaging: temperatures, CIP cycles and traceability data for HACCP and quality audits.
Cold chain
HACCP
Breweries & dairies
Smart home solutions
Thermostats, meters, air quality and security sensors across many homes: per-household portals for residents and fleet views for providers.
Comfort
Energy
Safety
How it works
Live in an afternoon, not a quarter
Four steps, no custom integration project.
01
Model
Create the tenant, describe your sites and assets — plants, parks, fields, networks or districts — and define device specifications with tags, units and limits.
02
Connect
Register devices and hand them an access key, point a connector at an existing API, or import historical files.
03
Monitor & act
Health checks, data-quality KPIs and your own rules run every minute. Alerts reach the right people in the portal, by e-mail and by push; operators respond from the control panel or the mobile app.
04
Analyse & share
Explore tags, build dashboards, schedule reports, share with colleagues, or query everything through the REST API.
Data quality
Data loss and bad readings are the real cost of IoT
A dashboard is only as good as the data behind it. The portal measures the completeness and sanity of every stream and tells you before your reports are wrong.
Offline detection that learns
Each device's reporting rate is measured from its own data; alerts fire after a few missed intervals, or after a fixed time you choose.
Coverage and loss per tag
Expected versus received points, gaps longer than twice the resolution, and downtime flagged by the monitor.
Anomalies explained
Values outside declared limits, statistical spikes, stuck sensors and counter resets, each listed with the reason.
Data quality · Line 3 · last 7 daysscore 85
Coverage
90%
Data loss
10%
Anomalies
7
Tag
Coverage
Anomalies
Score
temperature_c
99%
none
98
pressure_bar
93%
2 spikes
91
flow_m3h
71%
1 flatline · 3 h
64
energy_kwh
97%
1 counter reset
88
Pump station 3 stopped communicating.No data for 14 min; it normally reports every 5 min. Alert threshold 15 min.
Integrations
Open by default
Every capability is a documented REST endpoint with typed contracts, so the portal fits your systems instead of replacing them.
Device telemetry
Devices post batches of timestamped readings with a per-device key. Duplicates overwrite, cumulative counters are understood, notifications ride along.
HTTP connectors & files
Pull from any REST API on a schedule with time-window placeholders and any authentication scheme, or upload CSV / Excel history with schema detection.
REST API & agents
Aggregations, latest values, dashboards, reports and monitoring are all API calls with OpenAPI docs — ready for automation and the Pragma agentic framework.
Every tenant gets its own PostgreSQL and TimescaleDB databases, provisioned automatically. Data never shares a table.
Roles, custom roles and scopes
Six system roles, custom roles composed permission by permission, and access scopes that limit a member — or an AI agent — to specific assets and devices. Enforced by the API on every request.
Credentials handled properly
Per-device access keys stored as hashes, rotating refresh tokens in httpOnly cookies, connector secrets encrypted at rest.
Traceable
An audit log records who changed what, and invitation, activation and password flows go through verified e-mail.
Your data, your tenant, your rules
Sign in with your organisation's account, or ask your tenant administrator for an invitation. Platform administrators provision a new tenant in under a minute.