We build connected products end to end: embedded firmware, connectivity, ingestion and telemetry pipelines, device management, and the web and mobile apps people use to run it all. One team across the whole stack, so the integration problems that usually eat the timeline never become somebody else’s department.
Exelero builds connected products across the full stack: embedded firmware on the device, connectivity over Wi-Fi, cellular, BLE, LoRaWAN or Ethernet, cloud infrastructure for ingesting and storing telemetry at volume, device provisioning and over-the-air update systems, and the web dashboards and mobile apps operators use.
Typical projects include fleet and asset tracking, industrial and environmental monitoring, smart building and energy systems, and consumer connected devices. Engagements run as fixed-scope builds with handover, as a dedicated team alongside an in-house one, or as a discovery and architecture phase for teams deciding how to build.
The firmware team ships. The cloud team ships. The app team ships. Then nothing works together, because nobody owned the message format, the reconnection behaviour, the clock drift, or what the device does when the network disappears for six hours. Connected products are integration problems wearing a hardware costume, and the cost lands late — usually in the field, usually after the units are manufactured.
The second failure is scale. A pipeline that handles a hundred devices in a pilot is a different system from one handling fifty thousand, and the decisions that make the second possible are made in month one, not month ten.
Firmware on ARM Cortex-M, ESP32, STM32 and Linux-based targets; RTOS and bare-metal; sensor integration and calibration; power optimization for battery and energy-harvesting devices; secure boot and signed over-the-air updates; hardware bring-up and board support alongside your hardware partner.
Wi-Fi, BLE and Bluetooth Mesh, cellular including LTE-M and NB-IoT, LoRaWAN, Zigbee, Thread and Matter, Modbus and industrial fieldbus, MQTT and CoAP; offline-first behaviour and store-and-forward for intermittent links.
Device provisioning and identity, fleet management, telemetry ingestion at volume, time-series storage, stream processing, alerting and rules engines, digital twin models, edge computing where bandwidth or latency demands it, integration with AWS IoT, Azure IoT and Google Cloud IoT or a self-hosted stack.
Operator dashboards, mobile control apps with BLE pairing and provisioning flows, role-based multi-tenant access for installers and end customers, reporting and export, white-label deployments.
Device identity and certificate management, encrypted transport, secure key storage, penetration testing, and support for the regulatory work — CE and FCC processes, the EU Cyber Resilience Act, and sector-specific requirements.
Connected products are the exception on this site. Software — cloud, applications, device management — runs to our standard 8–12 week window. Firmware, hardware bring-up and any certification process do not compress the same way, so we scope IoT against a longer timeline rather than promise one we’d miss.
Everything else that makes us fast still applies: we start from foundations already built, scope is fixed in the first two weeks, one team covers the whole stack, and we decide together what to leave out of the first release.
Two to four weeks. Protocol and platform selection, data model, cost modelling at target fleet size, and a build plan. Useful on its own even if you build the rest elsewhere.
We build it and hand it over: source, infrastructure, documentation, and a walkthrough for whoever maintains it.
Our engineers inside your process alongside your people.
Firmware maintenance, fleet monitoring and feature development on a retainer.
Hardware startups who have a working prototype and need everything around it. Industrial and manufacturing companies instrumenting equipment they already own. Energy, utilities and smart building operators. Logistics and fleet businesses tracking assets. Agricultural and environmental monitoring. Established manufacturers adding connectivity to a product line that has never had it.
12–20 weeks from scope to launch. Software alone runs faster; firmware, hardware bring-up and certification are what extend it. Scoping gives you a firm date in week two rather than an optimistic one on day one.
We work on firmware, connectivity, cloud and applications, and alongside your hardware or contract manufacturer on board bring-up and integration. If you don’t have a hardware partner yet we can point you at ones we’ve worked with.
Yes, and it’s common. We start with an audit of the firmware, the pipeline and the cost model, and tell you plainly what’s salvageable.
Whichever fits your cost model and constraints — AWS IoT, Azure IoT, GCP or a self-hosted stack. At fleet scale the per-device cost differences are large enough to drive the decision, so we model it before choosing.
Signed, staged and rollback-capable, designed in from the start. Retrofitting OTA onto a fleet already in the field is one of the most expensive mistakes in connected products.
The architecture is chosen against your target fleet size, not your pilot size. Tell us where you expect to be in three years and we design for that.
Yes. Source, infrastructure and documentation on handover.
Bring the hardware, the use case, or just the problem. We’ll come back with an architecture, a scope and a timeline.