I build complete AI systems and infrastructure — from kernel-level Linux tuning to multi-agent orchestration in production.
Most "AI engineers" wire APIs together. I build the layer underneath: the operating system, the inference runtime, the orchestration logic, the observability.
End-to-end ownership of the stack.
struct AIEngineer {
name: &'static str,
role: &'static str,
location: &'static str,
focus: [&'static str; 5],
}
impl AIEngineer {
fn current() -> Self {
Self {
name: "Antonio Burgos",
role: "AI Systems Engineer · Creator of AurumOS",
location: "Palma de Mallorca, Spain",
focus: [
"AI infrastructure (C++23 / Rust / Python)",
"LLM serving (vLLM, SGLang, Ollama)",
"Multi-agent orchestration (Claude, MCP)",
"MLOps & observability (Langfuse, Phoenix)",
"Quantitative research (factor models)",
],
}
}
fn philosophy(&self) -> &'static str {
"If the tooling doesn't exist, I build it."
}
}The Linux distribution for AI engineers. Production-grade · macOS-Sequoia-class UX · adapts to your hardware.
A full Linux distribution. Built solo. ~50K LOC across C++23, Rust, QML, Python and Bash.
- 🖥️ Custom Qt6/QML desktop environment — 12 native apps + 3 Rust daemons over D-Bus
- ⚡ Forked Hyprland compositor tuned for GPU workloads (zero variance during training)
- 🎯 Hardware-tier adaptive — auto-detects GPU and adjusts every launcher (vLLM, Ollama, ComfyUI, fine-tuning scripts)
- 🧠 SOTA 2026 AI stack pre-wired: vLLM · Unsloth · DSPy · MCP · LiteLLM · Langfuse · Continue.dev · ComfyUI · YOLOv11 · SAM 2
- 📦 Steam-Workshop-style Model Pack Manager (6 curated bundles)
- 🧪 5-stage CI + 52-check smoke harness
- 🐧 Forked from Pop!_OS 24.04 LTS for NVIDIA / CUDA / cuDNN / TensorRT inheritance
➡️ github.com/AntonioBurgos91/aurum-os · ⭐ if it resonates
Auto-generated daily via self-hosted GitHub Actions — no external services involved.
| 🎯 Project | 📝 What it is | 🛠️ Stack |
|---|---|---|
| AurumOS ⭐ | Full Linux distribution for AI engineers · custom Qt6/QML desktop · SOTA 2026 AI stack · hardware-adaptive | C++23 Rust QML Python Bash |
| QuantumScore | Factor-based portfolio engine with backtesting · 5-pillar scoring · CAGR +101.4% / Sharpe 2.07 (validated strategy) | Python React Vite Backtesting |
| Inmobiliaria Palma | Real-estate finder · 5D scoring · 6K+ listings · undervaluation detection | Flask SQLite Leaflet Telegram API |
| AI Vision Assistant | Vision assistant powered by Claude API | Python Claude API CV |
| SOC Log Analyzer | Real-time security log analysis & threat detection | Python Security |
| Atlas GRC | ISO 27001 compliance & risk quantification platform | ISO 27001 Risk |
| 🏅 Certification | 🏛️ Institution |
|---|---|
| Google Data Analytics Professional Certificate | |
| IBM Data Science Professional Certificate | IBM |
| IBM SkillsBuild Analytics Certificate | IBM |
| Guided Learning Experience in AI | IBM |
| Google Cybersecurity Professional Certificate | |
| Google IT Automation with Python | |
| Curso Básico de Ciberseguridad | INCIBE |
mindmap
root((Antonio<br/>Burgos))
AI Systems
AI Infrastructure
LLM Serving
Inference Engines
Quantization
Multi-Agent
Orchestration
MCP Servers
Tool Routing
Agent Workflows
MLOps
Observability
Fine-tuning Pipelines
Model Serving
CI/CD for ML
Systems
Linux Distributions
Qt6/QML Desktop
Rust Daemons
Wayland Compositors
Quant Research
Factor Models
Backtesting
Portfolio Engines
Risk Analytics
"If a machine can do it better, a human shouldn't be doing it. If the tooling doesn't exist, I build it."
📍 Palma de Mallorca · Remote available · English B2
🌟 If you find my work useful, drop a ⭐ on AurumOS
