Experience

AKO Group · Head of Software

AKO designs and manufactures electronic controls for commercial refrigeration and the cold chain. When I joined, it was shifting from selling hardware to selling connected services through its IoT platform.

My role

I owned the end-to-end architecture and the technical decisions for all the company’s software; the team implemented them. Hands-on where it helped: data-processing scripts and working with developers at critical moments.

Scope

All software development in the company was under my responsibility.

• The IoT cloud platform: device communication, backend, data, client APIs and web frontend.
• The edge gateway software.
• Mobile apps: a field app to manage devices through a Bluetooth dongle, with a backend that let an installer connect remotely through the end user’s app to configure equipment on site; and an app to configure alarm systems through SMS commands.
• Legacy edge products, such as the standalone “Web Server” device, whose courtesy email relay for alarm notifications kept causing trouble: bursts of alarms from many customers went out through our account and hit sending limits. After several iterations, the final design held up well: sending moved behind an API with per-IP rate limiting, which smoothed out the peaks, and email ran from a separate AWS account, so a sending block could never affect the email of the main systems.

The platform

• Microservices on Node.js and Express, with Angular frontends built on a shared design system of reusable components.
• Device communication over MQTT with an in-house application protocol, and REST APIs.
• Traffic layer: introduced AWS WAF and an Elastic Load Balancer in front of nginx, which acts as reverse proxy to internal services, to secure the public endpoints by filtering malicious requests and balancing the load; and a CDN to cache selected content.
• Integration: a public API with webhooks so clients could extract data and receive alerts in their own systems; GraphQL piloted.
• Messaging: RabbitMQ queues between services, for example for pending reports and incoming metrics.
• Serverless where it fit: introduced AWS Lambda and API Gateway for coordination tasks and an in-house URL shortener that started as a proof of concept and reached production.
• Observability: introduced end-to-end tracing and metrics with OpenTelemetry and a self-hosted Grafana.
• Data: MongoDB for operational data and TimescaleDB for high-volume time series.
• Security by default: every API authenticated and validated against the user’s permissions, and database queries scoped to what the user is allowed to see.

Engineering practices

• Created the SRE function and built CI/CD from scratch with GitHub Actions and GitLab CI, with QA gates, Git workflows and Kanban.
• Introduced Architecture Decision Records (ADRs) and a technical roadmap.
• Introduced JMeter load testing to validate a service migration.

Stakeholders

• Product Managers for each electronic product family, each with their own requirements, later joined by business-oriented Product Managers focused on services. Balancing their competing priorities in a single platform roadmap was a core part of the role.
• Electronics and product design: mainly the firmware team, and the product designers who turned requirements into the specifications that reached electronics. Many decisions were made jointly and affected both sides.
• Manufacturing: I led an initial mapping of the business domain with strategic Domain-Driven Design. It was never completed, but along the way it led manufacturing to change industrialization processes and improve product traceability.
• An external software factory, which extended the internal team.

Responsibilities and results

• Platform architecture and technical roadmap; technology and vendor decisions.
• Scaled the platform from 15k to 32k+ connected devices.
• Technical lead for the software side of the €7M PERTE project: coordinating technical decisions across teams and bridging business and engineering.
• Grew the team from 2 to 5 internal engineers, and up to 14 people across development, systems and QA, including an external software factory.

Case studies: Scaling an IoT platform for the cold chain · Making an IoT edge gateway remotely maintainable

Independent Product Development

Two years building products of my own, fully hands-on.

• POLx: an interaction layer for live online events (event-owned chat, live polls and real-time layout control), built end to end with a partner using Vue.js and Node-RED and embedded in WordPress. Used at Superminds and by the Agència Catalana de Turisme, then acquired.
• Community platform for trainers: designed and developed a platform where trainers run their own student communities (Angular).

Case study: POLx, an interaction layer for live online events

Infoavisos (Grupo Intercom) · CTO

Incubated by Grupo Intercom as a spin-off of Tecnics, with both as partners, Infoavisos provided digital messaging for businesses: SMS first, then email marketing and mobile push notifications.

My role

Full technical direction with hands-on development, in a team of three developers, including me, at a small company where technology had to pay its own way.

Scope

• Enterprise SMS platform: mass sending with integrations for every kind of customer (HTTP API, email-to-SMS, FTP file drops, Excel and Access imports), evolved over 13 years while launching new products.
• Premium SMS services for clients such as Softonic, and SMS marketing campaigns for brands like IKEA.
• Montatuapp, a SaaS to build native iOS and Android apps from the web, built from zero.
• 50+ client projects across fashion, banking, energy, HR and automotive: mobile apps, intranets, access control and custom tools, including the first iOS and Android apps for iVoox.
• Early cloud adoption: moved company services to AWS in 2014 and introduced Docker for workload isolation in 2017.
• Hands-on code for years in ColdFusion (on the JVM), plus several AngularJS apps for clients.

Team and clients

Worked directly with the CEO and with clients. Set how the team worked: version control (SVN, then Git) and basic observability.

Case study: Montatuapp, native mobile apps configured from the web

Tecnics Enginyeria de Software · Senior Developer

A two-person software studio and my first job in software. The founder agreed the need with each client; I took the project from there: analysis, design, implementation and delivery, often at the client’s premises.

• Tick Tack Ticket (later acquired by Ticketmaster): maintained its point-of-sale system and designed its first access control system for major venues and events such as Palau Sant Jordi, FIB and Sónar; built a web chat before AJAX existed, with a desktop app for the operators.
• E-commerce and high-traffic web portals, from client requirements to production, including a business directory (herramientas.net).
• Mass SMS software: evolved the studio’s SMS sending software, with its own queue management. The product line that later became Infoavisos.
• A desktop backup application in Java for enterprise clients.
• A custom Webmin module in Perl, so event staff had business data and server administration in a single tool.