Meet Dan Sanabria, sharing his career highlights from being responsible for vintage-worthy telecommunications machines, to listening to 80s and 90s bootlegs in his day-to-day

What's your role here at WeBuild-AI?

I'm the Head of Platform Engineering at WeBuild-AI. In simple terms, my job is to make it as easy, fast and safe as possible for our teams to build and ship AI solutions. That plays out in two ways. Internally, I look after the shared platform, the tooling and the guardrails that our delivery and engineering teams rely on every day, so they can focus on solving the actual problem rather than wrestling with infrastructure. Externally, I am often hands on with client engagements, designing and running the cloud, MLOps and security foundations that our AI work sits on top of.

There is a rather nice twist to platform engineering in the AI era, which is that we are not only building the platforms that support AI agents, we are also working out how best to employ those agents ourselves. So one part of the team's focus is the shared foundations that let agents run safely, with the identity, memory, guardrails, governance and controlled tool access they need. The other part is putting those very same agents to work inside our own operations, taking on the diagnosis and the routine toil that platform teams have always carried. In other words, we are building the platform for everyone else and quietly building one for ourselves at the same time.

A lot of what I do sits at the meeting point of platform engineering, security and enablement. I care about building things that are reliable, well automated and secure by default, and about doing it in a way that other engineers can happily pick up and run with.

What does a typical day look like for you?

No two days are quite identical, which is one of the things I like most about the role. I usually start by catching up with the team, seeing where people are and clearing any blockers before the day properly gets going. A good chunk of my time then goes into architecture and design work, thinking through how a platform capability should fit together and reviewing designs with the team.

I try hard to stay hands on as well, so I am often in the code writing infrastructure as code with Terraform, building out CI/CD pipelines, or automating something that has been done manually one too many times. Mentoring runs through all of it. I am a firm believer in leading by example, so a lot of the coaching happens in the flow of the work rather than in a separate meeting. Security is never far away either, whether that is a threat modelling session, a review of how we manage identity and access, or a conversation with stakeholders about risk and priorities.

Tool-wise, the palette has grown quite a bit of late. Alongside the classics like Terraform, the major cloud platforms, Kubernetes, Git and the usual observability stack, I now spend a lot of my time with agentic frameworks such as Microsoft's Agent Framework, better known as MAF, and with vectorisation techniques and the vector databases that sit behind retrieval and memory for our AI systems. All that said, the most valuable tool is still honestly a good conversation with the right person at the right time.

What's your career journey been like so far?

My journey has been anything but a straight line, and I would not change that. I trained as an Electrical Engineer in Venezuela, and my first real love was big, complex telecommunications systems. Early in my career I worked on the kit that sat at the very heart of large mobile operators, gloriously ancient beasts by today's standards such as Lucent's 5ESS and Ericsson's AXE, along with the CDMA cellular networks around them.

To be fair to them, the versions I worked on were the latest incarnations of those platforms, properly state of the art for their day and the beating heart of national networks. The funny thing is how quickly state of the art becomes charmingly ancient. Judged by today's cloud-native standards they feel positively prehistoric, room sized machines that would make a modern smartphone weep, the wonderfully engineered and enormously heavy vinyl records of telecommunications. There was something quietly thrilling about being responsible for infrastructure that millions of people depended on without ever knowing it existed, even if some of that kit now belongs in a museum next to the fax machine.

I worked across network operations, teletraffic engineering and messaging platforms for operators in Europe and the Americas, and along the way I kept finding myself reaching for code to solve problems, writing small tools to make sense of data or to automate something repetitive. That instinct is really what pulled me from hardware and telecoms towards software. Over time I moved into automation, configuration management and what we would now call DevOps, working across the IT, media, government and financial sectors.

The more I automated, the more I cared about doing it safely, and that curiosity eventually led me to study for an MSc in Software and Systems Security at the University of Oxford. In truth, security had been a thread long before that formal step, and it has run through everything since, from securing cloud landing zones to baking controls directly into delivery pipelines. 

Rather than climbing a single ladder, I ended up wearing a lot of different hats along the way, often several at once and across many different projects. Depending on what a programme needed, I have been a SecOps lead, a solutions architect, and a platform engineering and DevOps specialist more times than I can easily count, including guiding a team of more than thirty engineers through a global digital transformation as the Head of Software Engineering for that programme. To me, platform engineering is where all of those threads really come together, which is what makes it feel like the natural home for someone who enjoys both the big picture and the fine detail.

Landing at WeBuild-AI and working on AI platforms feels like the latest chapter in the same story, taking everything I have learned about reliable, secure, well engineered systems and applying it to a genuinely exciting new field. Going from carrier grade telephone exchanges to agentic AI platforms in one career is not a bad way to measure how far the industry has travelled.

What do you enjoy the most about your job here?

What I enjoy most is the people and the pace of learning. AI is moving incredibly quickly, and there is a real energy that comes from working alongside people who are curious and generous with what they know. I love that I get to keep learning every single day, and that the person teaching me something new might just as easily be the most junior member of the team.

On top of that, platform work has a lovely sense of leverage to it. When you get it right, you make hundreds of other people's work easier, and seeing the team ship something they are proud of because the foundations simply worked is very hard to beat.

What's a piece of work that you'd be proud to show your past self, and why?

If I could show my past self one thing, it would be a data platform I helped build recently without writing a single line of code by hand. Rather than starting from a blank editor, we took a specification-driven approach, describing clearly and precisely what we wanted the platform to do and the constraints it had to respect, and then used agentic tooling to turn those specifications into working infrastructure.

The engineer who once spent years carefully hand crafting configuration for those old telecoms systems would find that almost unbelievable. What I am most proud of is not the novelty of it, but that the result was still well structured, secure and maintainable, because the thinking and the engineering discipline went into the specification rather than the typing. It is a good reminder that our job has never really been about writing code for its own sake. It has always been about solving problems clearly, and the tools we use to get there will keep changing.

Quick fire

How many languages do you speak?

Two human languages, Spanish and English, plus a fair few technical ones picked up over the years, from Bash and Python to Ruby and Terraform's HCL.

The Spanish word I wish English had is sobremesa, that lovely stretch of time when the meal is finished, but nobody wants to leave the table because the conversation is simply too good.

Coffee or tea, and go-to genre for deep focus work?

Coffee, without hesitation, and usually more of it than I would care to admit. 

For deep focus work I lean straight into 80s and 90s alternative, which is the music I genuinely love. I am also a rather serious collector of bootlegs from that era, so on a good day I am coding to a live recording that most people have never heard.

What's our company culture in 3 emojis?

🤝 🚀 🧠  (collaborative, ambitious, and endlessly curious)

Passionate about innovation?

Join a team that's architecting the AI generation of business.

Explore our open roles to see where you fit in, or follow us on LinkedIn to stay updated on new opportunities.