How Invenda Bridges Manufacturing & Digital with IoT

About Invenda

Invenda Group AG is a Swiss software company transforming automated retail and digital out-of-home (DOOH) advertising through its proprietary AI-powered platform. Founded in 2017, Invenda builds the connected software and hardware that brings e-commerce intelligence to physical retail. The company has deployed over 7,500 screens for clients including Mars Wrigley and Coca-Cola. As Invenda scaled, IoT software development became the core of its growth strategy: the automated retail platform, branded Luna, manages everything from remote pricing and real-time inventory tracking to programmatic ad delivery and predictive maintenance.

Invenda is not a traditional hardware manufacturer. The physical machines are one layer of a broader digital ecosystem. Every Invenda screen doubles as a media endpoint: the 49-inch touchscreens serve as both retail interfaces and programmable advertising displays. It is a model few competitors have matched, and it has positioned Invenda as a leader in global smart vending.

The Challenge: Scaling a Connected Retail Platform Across 22 Countries

As Invenda's fleet expanded across regions, managing it called for a more granular approach than the system had needed at its earlier scale. Configuring machines by location, assigning operator permissions across organizational levels, and adjusting pricing by market all needed to work at a scale only a few companies in the category have reached.

At the same time, the platform was generating data at a volume that called for infrastructure purpose-built to process it: transactions, ad impressions, sensor readings. The advertising system alone produced tens of millions of metadata records every few months, and the goal was to surface all of it as reporting operators and advertisers could act on directly.

As Invenda expanded into new markets and onboarded larger clients, a formal security assessment was scheduled as part of their ongoing program to ensure the platform met industry standards for protecting the data moving through the platform. All these workstreams were moving in parallel. The faster Invenda expanded, the more visible each need became.

How Accedia’s Team Embedded Across Workstreams

Accedia joined at a key moment in Invenda's growth, when the company was looking for additional engineering capacity to meet the demands of its expanding markets and a growing client base. Led by an Engineering Manager, Accedia's developers work alongside Invenda's architects and product leads across several connected areas: on-machine payments, cloud fleet management, digital advertising infrastructure, and security.

On-machine payment and transaction processing

Every vending machine in Invenda's fleet can run a different combination of payment terminals, sensors, and motor assemblies. Logic that works with one configuration may behave differently with another. Together with Invenda's engineering team, Accedia develops and extends the backend that accounts for these variations across the entire fleet: the transaction pipeline from basket to payment authorization to product dispensing. A separate hardware communication service manages interaction with physical components via the Multi-Drop Bus (MDB) protocol, handling motor retries in case products fail to dispense, payment terminal responses, and sensor monitoring.

Key features include partial refund logic (any item in a multi-item basket that is not dispensed is refunded automatically), remote power cycling from the cloud admin panel through Azure IoT Hub, and multi-price support that lets operators set different prices per payment method. MongoDB replicates data between the machine and the cloud bidirectionally. If connectivity drops, transactions work as normal, queue locally, and sync automatically once the connection returns.

Fleet management and digital advertising at scale

Luna introduces a hierarchical organization structure that lets operators manage thousands of machines by region, sub-region, and location. Accedia helps develop the platform, with role-based access control ensuring the right permissions at every level.

On the advertising side, the team helps develop the Atlas infrastructure, working on the integrations that connect Invenda's vending screens to third-party ad platforms. The screens handle several modes and report delivery data back to operators and advertisers. The network serves over 1 million ad plays every month.

Data analytics and reporting infrastructure

The machines generate transaction records, sensor readings, audience impressions, and ad performance metrics continuously across the fleet. The data pipelines run on Azure Databricks, where Accedia contributes to the ingestion, processing, and reporting layer, with machine learning workloads supporting predictive maintenance and user behavior analysis. The same infrastructure supports operational analytics: machine performance, locations needing attention, and regional variations in buying behavior. All data processing is designed to meet GDPR requirements.

Security assessment of payment and fleet APIs

Accedia's cybersecurity team assessed the Luna API and the Wallet API (Tapp) against OWASP Top 10 and OWASP IoT Top 10 vulnerabilities. Testing covered REST API endpoints, the Angular web interface, and MDB hardware communication protocols using Burp Suite Professional, Kali Linux, and directory enumeration tools. Identified gaps were addressed, and all industry-standard security measures have been implemented. For a company handling data at this scale across multiple markets, this assessment strengthens Invenda's position as a trusted partner to operators and the global brands that rely on its network.

Technology Stack

• On-machine backend: .NET/C#, MongoDB, gRPC, SignalR for transaction processing, hardware communication, and real-time machine-to-cloud sync
• Cloud and data: Azure IoT Hub, Azure Databricks, Azure Functions, PostgreSQL, Python for fleet configuration, data pipelines, ML workloads, and serverless event processing
• Frontend: Angular for operator admin panels and web interfaces
• DevOps and CI/CD: Azure DevOps, Azure Pipelines, GitHub for version control, continuous integration and deployment automation
• Cybersecurity: Burp Suite Professional, Kali Linux, gobuster for penetration testing and vulnerability assessment
• Protocols: MDB (Multi-Drop Bus) for low-level communication between vending machine components
• Security standards: OWASP Top 10, OWASP IoT Top 10

Where the Platform Is Heading

AI-assisted development and faster release cycles

Accedia and Invenda are bringing AI into the development process through GitHub Copilot, an AI-powered coding assistant that can help developers write code faster, catch errors earlier, and reduce time spent on repetitive tasks. To support broader adoption, the teams plan to consolidate their codebase and CI/CD pipelines by migrating from Azure DevOps to GitHub. They are also introducing feature flags so new functionality reaches production faster while remaining hidden from end users until it is ready to go live.

Turning automated retail into measurable value

Every layer described above exists for one reason: to make each location worth more to the people who depend on it.

For operators, that means precisely seeing what sells, where and when, and being able to act on it without standing in front of the machine. For brands, it means presence in places traditional retail cannot reach, with the transaction data to show what moved and why. For advertisers, it means a screen that reports what it delivered, in an environment where people stop, choose, and buy.

The work continues in that direction. More signal behind every decision, more of it available in real time, and less distance between what the data shows and the actions that follow.

Business Impact: How the Collaboration Accelerated Invenda's Growth

What began as a focused engineering engagement has become an expanding technical partnership. Since the collaboration began, Invenda's business has continued to scale. The company reports that operators using its automated retail ecosystem see up to 60% higher transaction revenue per machine and 30% fewer technical interventions. Fleet management, real-time telemetry, and IoT security are challenges that any organization building manufacturing IoT solutions will recognize.

Invenda is a software-first company. The technology is the product, and the machines are one way to deliver it. That focus is what has put the company at the front of its industry, and Accedia's engineers are helping build what comes next.