Machine Learning Engineer • Deep Reinforcement Learning • Control & Optimization

Machine Learning Engineer specialized in Deep Reinforcement Learning (PPO, SAC, TD3...) with experience in simulation-based environments and energy optimization.

Creator of Sinergym (⭐️ ), an open-source framework for applying RL to building energy control using EnergyPlus. Background in AI research with published work on RL evaluation and explainability.

Also interested in broader Machine Learning applications where RL, optimization, or decision-making systems can bring value.

• 🎓 PhD in Deep Reinforcement Learning (Artificial Intelligence) for Building Energy Optimization - University of Granada | 2021–2025
  • ⭐ Awarded with “Cum Laude” distinction (highest honors)
• 🎓 Master’s Degree in Computer Engineering – University of Granada | 2018–2019
• 🎓 Bachelor’s Degree in Computer Engineering – University of Granada | 2014–2018
  • ✔️ Specialization in Computing and Intelligent Systems

Reinforcement Learning

• PPO, SAC, TD3 and other deep reinforcement learning algorithms (Stable-Baselines3, RLlib)
• Continuous control and simulation-based learning environments
• Reward design, policy evaluation, and performance analysis

Machine Learning & Systems Design

• Design of ML systems for sequential decision-making problems
• Experimentation and evaluation of learning algorithms
• Modeling of simulation-based environments for RL research
• Engineering, MLOps and applicability

Tools & Frameworks

• PyTorch, pandas, numpy
• Gymnasium
• EnergyPlus
• Scientific Python ecosystem

Application Domains

• Energy optimization systems
• Control and decision-making under uncertainty
• Industrial simulation-based optimization

👉 https://github.com/ugr-sail/sinergym

An open-source framework for training and evaluating Reinforcement Learning agents in realistic building energy simulation environments.

• Integration of EnergyPlus + Gymnasium for realistic RL simulation
• Modular architecture for building and extending RL environments
• Benchmarking of deep reinforcement learning algorithms
• Standardized interface for continuous control problems
• Designed for reproducible RL experimentation
• Comparison with classical control strategies

• End-to-end RL workflows
• Containerized execution environments (Docker)
• Automated training and evaluation workflows
• Structured configuration-based experimentation
• Full documentation using Sphinx
• Unit testing and reproducibility practices

• Lead developer and system designer
• RL experimentation and evaluation methodology
• Architecture of simulation-based RL environments and project delivery (packaging, CI/CD, docs)

• 📄 Sinergym – A virtual testbed for building energy optimization with Reinforcement Learning (2025)
• 📄 Exploring Deep Reinforcement Learning Algorithms for Enhanced HVAC Control (2024)
• 📄 An experimental evaluation of deep reinforcement learning algorithms for HVAC control (2024)
• 📄 Explaining Deep Reinforcement Learning-Based Methods for Control of Building HVAC Systems (2023)
• 📄 Sinergym: a building simulation and control framework for training reinforcement learning agents (2021)

👉 Full list: https://scholar.google.com/citations?user=QDzMv44AAAAJ

• Applying RL to real-world systems
• Continuous Control
• End-to-end products (not only offline experiments)
• Energy & sustainability
• Industrial optimization
• Bridging research → production with engineering discipline (tests, CI, packaging, docs)

⭐️ If you're working on RL, control, or applied ML — feel free to connect!