Allenta Consulting is specialized in designing and implementing scalable solutions, with a strong focus on content acceleration, systems monitoring, observability, and security. Mumbo-jumbo aside, we solve complex problems for our customers as if we were creating the solution for ourselves, doing so efficiently, paying attention to detail, and reducing the amount of bs to the minimum.While my official title at Allenta is CTO, in a small company like ours, it doesn't mean much. I try to do my best on the projects I'm responsible for, being available to help my colleagues, and use the remaining time to continue the R&D approach started at dot2code.I do a lot of multitasking and exploratory work, but most of my time is spent on our partnership with Varnish Software. We provide support and professional services, extending the Varnish Enterprise product to solve customer needs, always collaborating with the awesome Varnish Software team. I'm particularly proud of the paywalling and traffic prioritization solutions we've built. They are running in major media and e-commerce companies in Europe and South America as well as during global events. This work involves mostly C programming, but also Python, Go, and a lot of VCL.In the monitoring area, we've also created products that I'm also proud of, such as a Zabbix to Grafana Cloud / Kafka / ... connector, and another similar one for the integration of VCS. We've also developed a Prometheus exporter to handle use cases important when coming from the Zabbix ecosystem. This work involves mostly Go, and there are other Python products that exist, but let's keep it short & sweet.In the R&D area, I'm proud of a rock-solid GELF Proxy implemented in C. We put a lot of effort into the performance and flexibility of the proxy, and the final result is awesome. I'm also proud of mundane but useful internal tools such as a Docker-based tool to streamline our development process, and an internal secret manager based on SSH keys.

dot2code Technologies was a small startup company owned by Allenta Consulting, a well-established consulting company, and it focused on creating great internet services. That's pretty generic, but that's exactly what we did: thanks to funding from the parent company and some boring cash-raising projects, we were able to explore and develop more exciting areas. In a sense, dot2code was the R&D side of Allenta.Our team consisted of 2-4 technical people, and we developed projects such as dot2code (a QR code tracking platform; RoR, MC, MySQL), eContact.me (a ticketing & social event management platform; PHP, MC, Redis, MySQL), a corporate URL management & tracking SaaS platform (AWS, Django, Celery, Python, Redis, MC, MySQL), Splunk extensions (e.g. Splunk STOMP Modular Input), Varnish Cache utilities and extensions (e.g. Varnish Bans Manager, VMODs, Varnish Event Logger, paywalling and traffic prioritization solutions, etc.; C, Python, Redis, Go), JIRA plugins (e.g. GitLab Listener, etc.; Java), Zabbix integrations (e.g. Zabbix Administration Console, Zabbix Plus, etc.; Python, PHP, Go, C), and more.The list of projects was long and diverse, with some being open-sourced, some never succeeding, and others unexpectedly succeeding. Some were just a way to learn and have fun, and a few of them are still alive and kicking as products commercialized by Allenta. My role in dot2code was to manage the technical side of the company, but due to the blurry line between dot2code and Allenta, and the success of some of the dot2code initiatives, I transitioned to a more general role as CTO of Allenta.

menus.es was a startup company located in Spain that provided a social platform focused on gastronomic experiences. With over 400,000 users in May 2013, the company had plans to expand to Mexico, Colombia, and Chile. Their platform included features such as coupons and discounts, a location-aware search engine, gastronomic blogging, and a gourmet club.While working at dot2code Technologies, I decided to join the company as a part-time consultant to help them evolve their previous old-fashioned architecture to the next level. I missed the good ol' days at Tuenti and I was looking for a part-time challenge. I provided general advice on many technical and non-technical questions and challenges during my short stay at menus.es. In particular, I introduced and set up a professional development environment (Vagrant), integrated a modern event logging and aggregation platform (Sentry), implemented an async task infrastructure (RabbitMQ + Redis), implemented app & systems level monitoring and reporting infrastructure, implemented professional site building and deployment procedures integrated with the base framework (Symfony 2), and introduced extra storage and caching layers based on Redis.Although it was a short stint, my time at menus.es was a valuable experience both technically and personally. However, the company wasn't prepared to invest in the necessary changes to move to the next level, so I made the decision to leave after a few months and return to my usual full-time job.

Tuenti was the website with the most traffic in Spain at the time, even surpassing Google according to 2010 reports by ComScore. With over 25,000 million monthly page-views and more than 500 servers, Tuenti was one of the most important services on the internet in Europe and counted almost 7 million users. In other words, Tuenti was an amazing place to work, especially after leaving the University of A Coruña - a difficult decision that I made after years of doubts.When I left my job at the University of A Coruña, I didn't have a plan, but the timing was perfect as someone at Tuenti contacted me due to a technical blog post I had written. After a few interviews, I was hired as a backend engineer responsible for designing and implementing an ejabberd-based (Erlang/OTP) chat backend routing 80-100 million daily messages. This was the company's first encounter with a technology like Erlang/OTP, and I was the only staff member with some knowledge of it. In other words, it was a huge challenge and a fantastic opportunity to learn and grow.I'm proud to say that the initial messaging backend I worked on was rock-solid and a success that continued evolving and growing for years. Once the chat was up and running, I worked on other non-Erlang/OTP components of the social network, such as the e-mail notification scheduling and delivery backend handling ~500,000 daily e-mails (PHP), the internal on-the-fly thumb generation service handling up to 5,000 reqs/sec (PHP, Google App Engine - Python), or the early design of the new photo storage/delivery architecture to handle the ~3 million daily uploaded photos (PHP).All of these were challenging projects given the scale of the social network, but none were as challenging as the Erlang/OTP-based messaging backend. Coupled with the fact that I never liked Madrid as a place to live in, I decided to quit Tuenti earlier than I initially expected and move back to A Coruña to start my own project with an old friend.

Thanks to Víctor M. Gulías (there's really no way to thank him enough, so let's keep it simple and bold), I was able to join the University of A Coruña as an assistant lecturer in 2005. This meant a formal five-year contract with the University, which was the first step towards becoming a full-time lecturer. More importantly, it meant that I no longer had to rely on random projects or grants to get paid, allowing me to focus on my research and teaching duties.As a teacher during this period at the University, I thoroughly enjoyed my role. I was responsible for the lab sessions (and a good portion of the theoretical classes) for the Information Systems Design course (4th term), which was essentially a part of the software engineering meta-course focused on high-quality engineering practices primarily based on GoF design patterns. Additionally, I was the director of several final-degree projects focused on Erlang/OTP, Ruby on Rails, J2ME & P2P technologies.As a researcher, I continued working on my PhD on P2P storage systems, Distributed Hash Tables, and other related topics. This included two research stays at the University of the West of England (Bristol, UK) and at the Swedish Institute of Computer Science (Kista, Sweden). However, I eventually grew bored of the research activities themselves and the dynamics of the academic world. I strongly missed the applied side of things, something beyond just producing research papers that solved problems that nobody cared about. After years of doubts, I decided to quit the University and join Tuenti.

The research side of the MADS (Models and Applications of Distributed Systems) group was always complemented by a strong applied side, both for funding reasons and because many team members had a strong applied background. One such project was the ARMISTICE (Advanced Risk Management Information System: Tracking Insurances, Claims and Exposures) project, which was a joint effort between the MADS group and the Risk Management Department at Inditex - the largest fashion group in the world. The goal was to develop a complex risk management information system on top of Erlang/OTP.During my time at the MADS group, in addition to my research duties and work on the VoDKA (Video on Demand Kernel Architecture) and e-MINDER (Electronic CoMmerce LeveragIng Network for Developing European Regions) projects, ARMISTICE was my main focus. I was responsible for analyzing, designing, and implementing the system, which was based on a distributed Erlang/OTP backend connected to a Java/Swing-based lightweight client using XML-RPC.While risk management may not be the most exciting topic in the world, the technical challenges of the project were fascinating, especially for a young software engineer like myself who was taking on such a significant responsibility for the first time. Although I eventually pivoted to other tasks in the research group, ARMISTICE was deployed in Inditex in early 2005, and I'm sure it helped the risk management team with their daily operations.

Working as a member of a research group like MADS (Models and Applications of Distributed Systems) in a public university is never easy. The group director, Victor M. Gulías, was always on the lookout for funding opportunities to keep the group going. Sometimes, this meant doing work that may not have been the most exciting, but it allowed group members to pay the bills and continue doing what they were truly passionate about.One such project was e-MINDER (Electronic CoMmerce LeveragIng Network for Developing European Regions), an EU Fifth RTD Framework Programme Project (IST-2000-28403) focused on developing a platform for creating e-commerce sites for small and medium enterprises (SMEs) in the European Union.In the e-MINDER project, I worked as an analyst, responsible for developing teaching materials on topics like e-commerce basics, security, logistics, and marketing, as well as providing training to SMEs. This allowed me to continue my research activities with the group and my collaboration with the VoDKA (Video on Demand Kernel Architecture) project as a software engineer.

The MADS (Models and Applications of Distributed Systems) Group is a research group at the University of A Coruña in Spain that specializes in designing and implementing distributed systems. The group is affiliated with the Computer Science Department and was founded and led by Professor Víctor M. Gulías. Thanks to Víctor, I was able to join the group as a software engineer immediately after completing my Master's degree project, which was also supervised by him.This position was my first formal role during my eight years at the University of A Coruña. During this time, I was primarily involved in the VoDKA (Video on Demand Kernel Architecture) research project. Essentially, it was a visionary -that's the right word to describe a genius like Víctor- and early version of Netflix built on Erlang/OTP, a unique platform that was considered exotic at the time. This project provided my first real-life experience with Erlang/OTP, and it was an excellent opportunity for me to learn and develop as a junior software engineer.These were the early days of VoDKA as a product, and I was able to contribute to the design and implementation of the storage, caching, and streaming layers of the streaming server. It was also a great opportunity for me to learn about the challenges of building a real-life distributed system like this, which was later commercialized, extended, and deployed by the group's spin-off, LambdaStream.