Data usage dropped from 3–5GB to under 300MB per month, cutting costs and easing low-signal connectivity strain.
EV charging, billing, invoicing, alerting, compliance, and ERP sync now run as one system.
After 14 straight reports at Amber, the project has held Green status for 5 consecutive months.
The client provides energy management technology for multi-occupant buildings, monitoring and controlling electricity, EV charging, and solar production through a custom hardware gateway. As their platform scaled past 25,000 connected devices through more than 1,700 custom gateways, growing data volumes strained low-signal connections, and delivery was split across rotating freelancers with no single owner or roadmap, prompting the search for a stable, accountable partner to support their next stage of growth.
Gateways were transmitting 3–5GB of data per month over limited 4G connections at remote sites, making reliable syncing difficult and driving up transmission costs.
Offline periods often led to gaps in recorded data and incomplete syncing once connectivity returned, impacting data accuracy and user confidence.
An outdated Python environment was blocking new feature development, while inconsistent device management left services running unnecessarily and increased the risk of crashes.
The local web interface made day-to-day device management difficult for field technicians, slowing down routine operations and troubleshooting.
Work had been distributed across multiple contractors without a single accountable owner or shared tracking system, an opportunity to bring structure, consistency, and a clear roadmap to ongoing development.
Core systems like the Mobility module and address/data model needed to be rearchitected at the same time as unfinished work from a departing contractor had to be completed — all without slowing the rest of the roadmap.
Rebuilding the Platform for Leaner, More Reliable Data Handling
Glorium Technologies consolidated delivery under one accountable team and rebuilt core parts of the platform for reliability and scale, cutting data usage, strengthening offline handling, modernizing the underlying stack, and improving the field experience for technicians.
Work that had been spread across a rotating group of freelance contractors was brought under a single dedicated PM and a stable three-person core team. Project tracking was consolidated onto one shared system (Linear), and reporting moved from ad hoc updates to a structured bi-weekly, then monthly, cadence, giving the client consistent visibility into progress.
The team introduced local event-based data processing to sharply cut data volume, migrated parts of the datastore from MongoDB to SQLite for better stability, and modernized the underlying Python environment to unblock new feature development. A Dockerized local development setup was also introduced to make ongoing work more consistent across the team.
Beyond stability work, our team rebuilt and expanded major functional areas: OCPP-based EV charging control (remote start/stop/reset, a WebSocket gateway, and a command dashboard), a rebuilt alerting system with multi-channel notifications and escalation logic, and a reworked billing and invoicing engine with payment-gateway integration and financial reporting.
We integrated the ISO-based Swiss data exchange standards used by utilities and building accounting systems, and connected the platform to the client’s wider ERP ecosystem (Salesforce CRM, a custom ERP, in-house accounting systems, and local solutions like Bexio and Abacus) for ERP and invoicing sync.
Consistent Delivery and a Platform That Keeps Up With Growth
Local event-based processing replaced constant raw data streaming, reducing usage from 3–5GB to under 300MB per month. This lowered transmission costs and improved sync reliability across sites with limited 4G connectivity.
The system now manages EV charging control, billing and invoicing, alerting, compliance reporting, and ERP sync across 18 active workstreams, 25,000+ connected devices, and 1,700+ gateways — supporting the client’s expansion into new markets.
Migrating parts of the datastore from MongoDB to SQLite and modernizing the underlying environment reduced crash risk and unnecessary system load, creating a more stable foundation for ongoing feature development.
Work that had cycled through 13 different contractors was consolidated under one dedicated PM and a stable three-person core team, with a single shared tracking system and a consistent reporting cadence. Project health moved from Amber for 14 consecutive reports to Green for 5 consecutive months.
Integration with Swiss data exchange standards, along with sync across Salesforce CRM, Bexio, and other ERP and accounting systems, brought the platform in line with regulatory requirements while streamlining invoicing and financial reporting.
Powering an Energy Management Platform at Scale
| Layer | Technologies | Rationale |
|---|---|---|
| Backend | Python, Django | Powers the core platform logic — billing, alerting, tariffs, and reporting — with a framework mature enough to support rapid feature expansion. |
| Edge / Gateway | Raspberry Pi, Linux | Runs on-site at each building, handling local data collection and event-based preprocessing before anything is transmitted. |
| Data Storage | MongoDB, SQLite | MongoDB serves the broader platform, while SQLite handles lightweight, local storage on gateway devices for better stability and performance. |
| Device Communication | MQTT, OCPP | MQTT handles lightweight messaging between devices and the platform; OCPP manages EV charger control — remote start, stop, and reset. |
| Integrations | Salesforce, Custom ERP, Bexio, Abacus | Syncs billing, customer, and accounting data across the client’s ERP ecosystem, including Salesforce CRM, a custom ERP, in-house accounting systems, and local tools like Bexio and Abacus. |
| Observability | Sentry | Monitors system health in production, surfacing issues before they affect uptime. |
| Dev Environment | Docker | Provides a consistent, containerized local development setup across the team. |
If you’re responsible for a hardware-connected platform — IoT devices, gateways, or sensors — you already know that reliability doesn’t come easy when your sites run on limited or unstable connectivity.
As you scale across hundreds or thousands of devices and serve different types of users, the data and architecture decisions that worked early on stop holding up. You may also recognize a different kind of strain: delivery that’s become fragmented, spread across contractors without a single owner or shared roadmap. If you’re rebuilding a platform that’s outgrown its architecture, or looking to bring delivery under one accountable team, the approach in this case, covering data optimization, delivery structure, and compliance, was built for exactly that situation.
Multi-Role Platforms
Regulatory Compliance
IoT & Energy Tech
Low-Connectivity Environments
Device Fleets
Delivery Consolidation
Book a free intro call, bring your current challenges, and we’ll walk you through how the same approach could apply to your platform.
