I design software, orchestrate AI agents, and build the context that connects them. Architecture, bounded execution, and verifiable outcomes, not code generation alone.
I use AI agents to explore, implement, and review code. I experiment with new ways to design and develop software, building systems that coordinate agents. Autonomy accelerates production; engineering judgment keeps it under control. I use AI agents to explore, implement, and review code. I experiment with new ways to design and develop software, building systems that coordinate agents. Autonomy accelerates production; engineering judgment keeps it under control.
context.md
Context
Explore the system, capture the intent, and define the constraints. Useful agents need useful context: architecture, dependencies, acceptance criteria, and clear boundaries.
Read the code and its dependencies
Define the goal and constraints
Break the work into bounded tasks
tasks.md
Orchestrate
Coordinate agents to explore, implement, and review solutions. Distribute responsibilities and maintain a system-wide view without losing control over architectural decisions.
Delegate tasks with defined responsibilities and boundaries.
Coordinate parallel execution.
Integrate small, coherent, reviewable changes.
review.checklist
Verify
Review each result against the defined objectives. Read the code and check its behavior before integrating changes. Execution can be delegated, but technical responsibility cannot.
Fraud detection system based on an event-driven architecture with Spring Boot microservices and Kafka messaging, instrumented with metrics, distributed tracing, and Grafana dashboards.
Agents expand what I can build. Engineering judgment determines what is worth building and how to make it reliable.
How I work
My work connects systems engineering with agentic development. I design the architecture, shape the context, and use agents to explore and implement within clear boundaries. The goal is not more generated code: it is software that can be understood, verified, and evolved.
What I study
I am pursuing a Bachelor’s degree in Computer Science at Universitat Politècnica de València. This education allows me to delve deeper into the foundations of computing, understand systems from first principles, and develop a broader perspective on software engineering.
What I’m looking for
I want to work with teams that take architecture seriously, explore new ways to use AI agents, and make engineering judgment a core principle. Teams that use these new capabilities to solve real problems without compromising quality, responsibility, or control over technical decisions.
Engineering beyond generation.
Agentic development makes context, architecture, and verification more important, not less. These are the foundations I bring to the workflow.
Context & orchestration
Clear goals, bounded tasks, coordinated execution
Give agents the context they need, define task boundaries, and coordinate work around explicit goals. Keep decisions and integration under human ownership.
Context engineering
Agent orchestration
Task boundaries
Human review
Leverage without losing direction
Architecture & systems
Domain boundaries, events, processes, memory
Understand the system underneath the abstractions. Clear domains and interfaces make software easier to reason about for people and for the agents working with them.
Java
Spring Boot
DDD
Hexagonal
C
POSIX
Structure that survives change
Verification & delivery
Tests, observability, reproducible environments
Generated code is a proposal. Tests, review, and observable behavior turn it into a result that can be trusted, shipped, and operated.
Testing
Observability
Grafana
Docker
Nix
CI/CD
Evidence, not plausible output
Professional experience
Working on real projects has taught me that software development goes far beyond writing code. It means understanding existing systems, making informed technical decisions, and building solutions that can be maintained and evolved over time.
May 2026 - PresentSoftware EngineerCapgeminiDetails
Software Engineer working on an enterprise solution for the insurance industry. Part of an international, English-speaking team delivering software for a global client.
Develop and maintain features for a large-scale enterprise insurance platform, working primarily with Java 21.
Analyse technical requirements and existing system behaviour to design and implement changes within a mature production codebase.
Investigate and resolve application issues by tracing execution flows, business logic and configuration dependencies across multiple components.
Collaborate daily with developers, analysts and stakeholders across international, English-speaking teams within an Agile/Scrum environment.
Jun 2024 - Feb 2026Full-stack Software DeveloperDigital Value S.L.Details
Full-stack development in real production environments, including a booking system, an IoT management platform, and a headless CMS that played a key role in a complete municipal web portal renovation.
Designed and developed a headless CMS for a virtual newspaper archive, including data model migration and editorial workflow improvements.
Developed an IoT device management platform with Node.js, supporting hundreds of concurrent devices with real-time monitoring and remote control.
Implemented end-to-end observability in a reservation system connected to IoT relays for caravan power supply.
TypeScript
Node.js
REST APIs
IoT
Observability
Grafana
CI/CD
Mar 2023 - Jun 2023Frontend DeveloperVinfoVALDetails
Developed and maintained the Angular front end of a police management PWA, optimized for mobile use, used by law enforcment agencies to consult and manage operational information during real-time interventions.
Developed Angular interfaces optimized for mobile use, prioritizing usability and data access speed in the field.
Refactored critical components, reducing technical debt and improving front-end maintainability.
Integrated REST APIs in collaboration with the backend team and participated in code reviews and CI/CD workflows.
Front-end architecture
SOLID
TypeScript
Angular
PWA
CI/CD
Code reviews
Mar 2022 - May 2022Atlassian Consultant / Full-stack DeveloperInlogiqDetails
Training and hands-on work with the Atlassian ecosystem, focused on advanced Jira configuration, process automation, and custom plugin development.
Performed advanced Jira and Jira ITSM configuration, adapting workflows, permissions, and schemes to project requirements.
Resolved complex Jira incidents through automations, Groovy scripting, and marketplace plugins.
Developed an in-house plugin for Jira Software using Java and Spring Boot, including web interface support and custom business logic.
Terminal-first, reproducible, and built for agent-assisted development. Orchestration tools, editors, and infrastructure belong to one engineering workflow.
Agents & AI
Orca Multi-agent orchestration
Agent development environment for multi-agent orchestration.
ChatGPT Development assistant / LLM
Advanced debugging, PR reviews, system design, and help with technical decisions.
GitHub Copilot Pair programmer
Intelligent autocomplete, test generation, and refactors inside the IDE.
Opencode Code agent
Code agent integrated in the terminal, connected to GPT 5.6 Sol as the main LLM, to automate writing, refactoring, and code analysis within the development flow.
Terminal
Neovim Primary editor
With Nixvim, native LSP, telescope, treesitter. No unnecessary plugins.
kitty Terminal emulator
GPU-accelerated terminal with panes, tabs and extensive customization.
fish + starship Shell
User-friendly shell with fast prompt, git awareness and rich completions.
ripgrep + fd + fzf Search
Replacement for grep/find. Integrated with Neovim and custom scripts.
zoxide Directory navigation
Smart cd replacement to quickly jump to frequent directories with short commands.
atuin Shell history
Interactive shell history with fuzzy search and synchronization across machines.
lazygit Git TUI
For complex git operations. Interactive staging, visual rebase.
yazi File explorer
Fast terminal file manager with previews, vim-style navigation and scripting support.
bat + eza File viewing & listing
Modern cat and ls replacements with syntax highlighting, icons and git integration.
Desktop
Quickshell Desktop shell
I use Caelestia's config as a base and adapt it over time to match my daily workflow.
IntelliJ IDEA Backend IDE
For large Java projects when I need deep navigation, debugging, and advanced Spring integration.
Firefox Web browser
Primary browser for daily work, development, and debugging with strong privacy defaults.
Bruno API testing
Text-based API client, ideal for testing REST endpoints, organizing project collections, and versioning them with git.
System
NixOS Operating system
Fully declarative config. Flakes for total reproducibility.
Home Manager Dotfiles
All user config managed with Nix. One repo, any machine.
Hyprland Wayland compositor
Dynamic tiling, minimal setup and smooth animations on Wayland.
Docker + Docker Compose Containers
Containers and local orchestration of services for development and testing in reproducible environments.
No cookies, no IPs, no fingerprinting. A random token in your browser and the approximate city Cloudflare's edge already resolves, rounded to ~10 km.
Available for new opportunities
I'm interested in engineering teams that combine strong systems thinking with agentic ways of working. Real technical challenges, clear ownership, and software built to be understood, not just generated.
