Guillermo Madariaga Sierra
Systems · Architecture · 2026
Five years sustaining two products in production and carrying them from requirement to release. This portfolio is built with the same rules I apply to them.
Background
Electronics taught me to follow a signal all the way through and to respect the whole system. Software became the medium for applying that way of thinking to products that have to last.
I have been at Speed Solutions for five years, working on two products in production: SpeedFlowLink, a multi-tenant cloud platform, and SpeedPrintLINK, an industrial Android application. I built the centralization integration that connects them.
My involvement does not stop at implementation. I gather the requirement with whoever raised it, design the solution, build it, test it, integrate it and see it through to release — across backend, front end and the data model.
Track record
From an electronics degree to end-to-end product work. Every milestone left something I still use.
Education
Electronics engineering. The final project worked with neural networks and fuzzy systems, which left me a concrete sense of what sits underneath the AI tooling I use today.
Outside work
I build tools for the problems I run into: remote diagnostics, traffic capture, device monitoring. Several started as something narrow and grew into complete products, with an architecture and tests of their own.
Systems in production
Software other people use every day to do their job. Here decisions are not measured in elegance but in what it costs to change them two years later.
Multi-tenant cloud platform / Production
Cloud platform for managing fuel dispensing stations — a Blazor front end, several APIs with separate responsibilities, and a shared multi-tenant data model.
Role: Backend, front end, data model and integrations
Industrial Android application / Production
Android application for custody transfer control at dispensing stations, with a local database, hardware integration and synchronization with the cloud platform.
Role: Maintenance, evolution, integration and testing
Own products
Each one started because I needed something that did not exist, or existed halfway. The ones here are those that grew an architecture of their own.
Diagnostics platform / Own product
Distributed logging platform with an API, a client, shared contracts and a dashboard, born from a real field-diagnostics problem.
Developer tooling / Windows
A modular reverse proxy that exposes local applications to the network and captures, persists and inspects every HTTP exchange passing through it.
Own products
The next two live where software meets hardware: a protocol between two surfaces, and a bridge between two transports.
Server + client system / Multi-surface
A remote control system made of a Windows server, a mobile client and a shared message model that defines the protocol between them.
Desktop monitoring / Serial and TCP
A desktop tool for watching serial and TCP logs in real time, with an integrated terminal for configuring the device without switching applications.
More work
Smaller projects, nearly all born of a specific need. They are listed because together they say something none of them says alone.
How I work
Enough structure to protect what matters, no ceremony out of habit. Every decision in this portfolio comes with the alternative I rejected.
A domain that does not know who calls it can be tested, moved and sustained. Everything else is a replaceable detail.
Every abstraction is paid for in future readings. I add one only after the problem has shown up twice.
Cross-cutting rules in one place, decisions written down, migrations reviewed. What is scattered never gets audited.
Delivery
I do not hand off a ticket — I follow a change until it is in production and someone is using it.
Architecture
The domain should not know whether a browser, a field device or a background process is calling it. Ports express intent; adapters handle the outside world.
What this site claims I can do is done in the site itself: a domain with no framework dependencies, use cases separated from adapters, typed content, unit and end-to-end tests, continuous integration, and a small service with a database behind the contact form.
See the codeStack
Specialized in .NET, with enough range that an idea reaches production without depending on someone else to cover the gap.
Backend
Data
Frontend
Mobile & desktop
Practice
The delivery and AI tooling I use daily. The choice matters less here than the discipline behind it.
Delivery
Applied AI
Applied AI
I use agents, MCPs and skills on real projects, from gathering the requirement to the release. My degree covered neural networks and fuzzy systems, which gives me a concrete sense of what sits underneath. I supervise assisted development directly in the editor and the terminal, and ownership of what ships stays with me.
Contact
A product, a system or a technical problem is a good place to start.