Pushing limits driven by curiosity.
I build the engineering organisations behind complex systems people depend on — today, the Earth intelligence platform at SkyFi.
Engineering leader with a strong bias toward technology, an open mind, and genuine curiosity about how things work.
As CTO at SkyFi I lead the engineering organisation building an Earth intelligence platform — making satellite imagery and geospatial analytics accessible to people who have never touched a GIS tool. That spans a commercial marketplace, AI products, government and defence programmes, and the infrastructure underneath all of it.
Twenty-plus years in, I still start from the customer’s problem rather than the tooling. The right technology is the one that answers the question they actually have, not the one the team would most enjoy building with. That is why the list below keeps growing, most recently with AI.
I wrote these as SkyFi’s engineering principles. They decide what we build as much as how we build it, and none of them are new — they are how I have worked for most of my career. Putting them in writing changed one thing: it made them something the team can hold me to.
You build it, you run it.
Responsibility does not end at the merge, or at the release note. It ends when the thing is in someone’s hands and solving the problem it was built for. That means caring about the shape of the feature as much as the reliability of the service behind it — and fixing what you find broken without waiting for permission.
Don’t wait to be told.
The best teams see the problem before it is filed: the customer workflow about to break, the market moving, the root cause that will cost ten times more next quarter. Bring the idea as well as the fix. When in doubt, do something — stagnation usually costs more than a wrong decision you can correct.
Respond to change over following a plan.
Ship small, ship often, and let real usage settle arguments a roadmap document never will. Don’t fall in love with your first solution: the plan is a hypothesis and the people using the product are the ones who get to falsify it. Maximising the work not done is a genuine skill, and an underrated one.
Speak truth to power.
Someone has to say the roadmap is wrong, the estimate is fiction, or the feature nobody uses should be deleted. Challenging a plan is as valuable as challenging an architecture, and usually less popular. It only happens when failure is survivable, which makes psychological safety a precondition rather than a perk.
Speed is a feature.
Speed is how fast you learn, and it is also something customers feel directly. Take the friction out of the path from idea to production, invest in the pipeline until you trust it, and remember most decisions are reversible — a good decision today beats a perfect one next quarter.
Augment your intelligence.
Two questions, always: how does AI make this problem smaller for the team, and how does it make the product genuinely better for the person using it? Both matter, and they are not the same question. Stay current, experiment in the open, and stay honest about what the tools cannot yet do.
Jul 2022 — present · 4 yrs 2 mos
SkyFi is democratising access to satellite imagery. I lead the engineering organisation — around 45 people across five teams covering the commercial platform, AI, government and defence programmes, and the infrastructure underneath all of it.
Own technology strategy and the engineering organisation end to end — commercial platform, AI products, government programmes and infrastructure.
Led the engineering team building the next generation Earth observation marketplace and platform for data and analytics.
Built out the engineering team and delivery practice as SkyFi scaled from early product to a platform business.
Oct 2020 — Jun 2022 · 1 yr 9 mos
Built innovative software for medical diagnostics — advancing imaging-based diagnosis and personalised treatment to improve human health and wellbeing.
Career consulting and advisory for engineers navigating the step from individual contributor into senior and leadership roles.
Dec 2019 — Jul 2020 · 8 mos
Led the team through the integration into Mirantis following the Docker Enterprise acquisition — holding delivery and morale steady across an ownership change.
Jun 2019 — Dec 2019 · 7 mos
Responsible for building a team from scratch in Sofia while leading Docker Hub (~12 engineers) and Docker DTR (6 engineers) remotely.
The Sofia site onboarded UX designers, QA automation, infrastructure and security engineers. We hired ten engineers and shipped features across Docker Hub and DTR, including 2FA for Docker Hub.
Jul 2017 — May 2019 · 1 yr 11 mos
Led four teams (one remote) across shared services, platform and IaaS — responsible for delivering high-quality PaaS and SDLC tooling to product teams, including the containerisation effort that moved services onto Docker images and Kubernetes.
Organised the company’s first internal hackathon: 50 engineers, 12 teams, roughly 30% of the IT workforce. It produced two shipped features and two prototypes that went into the backlog.
Apr 2014 — Mar 2017 · 3 yrs
Grew the site from zero to 50 engineers, onboarding two managers and twenty-plus engineers, with full ownership of the domains the site held.
Led work end to end — requirements through rollout and support — across partner banking details management, KYC identity verification, automated receipts, invoicing on behalf of partners, and real-time tax calculation and reporting.
Oct 2007 — Apr 2014 · 6 yrs 7 mos
Delivered user interfaces across the vSphere platform, managing teams of 10–20 engineers directly and 40+ indirectly on a single programme.
Led the multi-year, multi-site effort that moved the vSphere management interface from a C# desktop application to a Java/Flex web application, delivered in 2013, and presented the vSphere Web Client at VMworld US 2012.
Mar 2003 — Sep 2007 · 4 yrs 7 mos
Design specifications, delivery, team leadership, task management and mentorship of junior engineers across a range of engineering projects.
Grew the engineering organisation to ~45 engineers across five teams — commercial platform, AI, government and defence, and infrastructure — taking it from early-stage product work through the company’s Series A and into programme work with NATO and U.S. defence customers.
Phase I prototype for U.S. Special Operations Command: mission-critical geospatial imagery delivered to the tactical edge, with an integrated Android Tactical Assault Kit plugin.
Built from zero to 50 engineers and two managers, holding full ownership of partner banking and payments, KYC identity verification, invoicing on behalf of partners, and real-time tax calculation.
Stood up a second site from scratch — ten engineers spanning UX, QA automation, infrastructure and security — while leading Docker Hub and DTR remotely. Shipped 2FA for Docker Hub.
The multi-year, multi-site migration of VMware’s management interface from a C# desktop application to a Java/Flex web client. Delivered 2013; presented at VMworld US 2012.
Four teams across shared services, platform and IaaS, delivering PaaS and SDLC tooling and moving services onto Docker images and Kubernetes.
SkyFi among five companies advanced from a cohort of 150, moving toward operational adoption across the Alliance.
Selected for Phase I development of a Sovereign Intelligence Platform prototype and ATAK plugin.
Invited among 250 Bulgarian founders, investors and operators from around the world.
Presented the vSphere Web Client, VMware’s browser-based management interface.
I’m happily building at SkyFi and not looking for a role. I am always glad to talk about advisory and board work, speaking, and comparing notes with other engineering leaders — particularly anyone working on Earth observation, defence tech, or scaling a team through the awkward middle.