Álvaro Jiménez Martínez

Málaga, Spain

Backend Engineer specializing in Applied AI & LLM Integration, high-performance distributed architectures, and Web3.

Skills & Technologies

Some technologies I've been working with:

AI & LLMs Systems Engineering

Core Focus

Autonomous agents (LangGraph, pydantic-ai), QLoRA & Apple Silicon MLX fine-tuning, AI safety guardrails, and Vision-Language models.

50% VRAM Reduction (mlx-modernbert)
10/10 Honor Thesis (3DGS + SAM 2)
Python (FastAPI, PyTorch) LangGraph & pydantic-ai QLoRA & Apple MLX Guardrails & Safety SAM 2 & OpenCLIP OpenCV & 3DGS

Backend & Web3

High-throughput FastAPI/Node.js microservices, PostgreSQL semaphores & Polygon blockchain assets.

€2.2M+ Polygon Assets
PostgreSQL Semaphores
Node.js PostgreSQL React & WebGL Web3.js & Polygon Java

DevOps & Automation

Automated n8n Jira incident workflows, AWS CI/CD pipelines, Docker, and observability.

553 Tickets Resolved (Azure LLMs)
AWS CI/CD & Coolify
Docker & AWS CI/CD n8n Workflows Prometheus/Grafana Playwright

Education & Honors

B.Sc. in Software Engineering — University of Malaga (GPA 8.6 / 10.00 — Top 10% of Class)

EN: ProfessionalES: Native
Local LLM Fine-Tuning & Coding Agent Harness
QLoRA Fine-Tuning
Python QLoRA MLX LangGraph LLMs Subagents

Local LLM Fine-Tuning & Coding Agent Harness

AI Systems Engineering project optimizing a local coding agent harness with subagent-driven development, custom skills, guardrails, and fine-tuning Gemma models with QLoRA on Apple Silicon.

Semantic 3D Farm Reconstruction
Grade: 10/10 (Honor)
FastAPI PyTorch 3D Gaussian Splatting SAM 2 OpenCLIP React WebGL

Semantic 3D Farm Reconstruction

Bachelor's Thesis (Grade: 10/10 Honor) — Distributed precision agriculture platform leveraging 3D Gaussian Splatting and Vision-Language Models (SAM 2, OpenCLIP) to generate semantically segmented 3D crop models.

Aegis: AI-Powered Browser Automation Agent
Accessibility Tree Engine
Python Playwright pydantic-ai LLMs Agentic Workflows

Aegis: AI-Powered Browser Automation Agent

Autonomous browser agent utilizing Chromium native Accessibility Tree as a semantic perception engine (Python, Playwright, pydantic-ai) with dynamic token optimization.

mlx-modernbert
Reduced VRAM by 50% via fp16
AI Backend MLX Apple Silicon Python NLP

mlx-modernbert

Pure MLX training framework for ModernBERT on Apple Silicon, optimized for M-series neural accelerators.

My Trajectory

My studies and professional milestones.

Sep 2022 - Jun 2026

B.Sc. in Software Engineering at University of Malaga

Software Engineering Degree with an 8.6 / 10.00 GPA (Top 10% of Class). Relevant coursework: Computer Vision, Deep Learning, Distributed Systems, Cloud Computing, and Software Architecture.

May 2025 - Dec 2025

Full Stack Web3 Developer at UnknownGravity

Sole backend developer for a real estate tokenization startup. Led end-to-end architecture on Polygon launching a marketplace with over €2.2M in listed property assets (Node.js, PostgreSQL, Web3.js, React, AWS, Docker, GitHub Actions).

Mar 2026 - Jun 2026

AI Engineering Intern at BHS Corrugated Spain

Industrial AI & Automation. Designed safe-by-design n8n/Python AI workflows to automate Jira incident diagnosis for corrugator machines, reducing manual handling by 40% and resolving 553 tickets automatically via Azure LLMs.

Jun 2026 - Present

AI Systems Engineering & Research

Developing native Apple Silicon MLX training frameworks (mlx-modernbert), fine-tuning local Gemma models with QLoRA, building autonomous browser agents (Aegis with pydantic-ai/Playwright), and subagent-driven coding harnesses.