10/13/2026 updated
100 % available
AI Research Engineer | Deep Reinforcement Learning & Energy Systems
Pisa, Italy
Worldwide
PhD in Sustainable Development and Climate Change with specilization in AI and EnergyAbout me
I'm a Machine learning Engineer at Imperial College London. My work bridges machine learning, energy systems, and data-driven sustainability from smart grids to renewable energy communities and EV integration. Tools: Python, R, Julia, TensorFlow, PyTorch, SQL, AWS, Scikit-learn, Keras, and Power BI.
Python (Programming Language)Machine LearningMicrosoft Power BIAzure Machine LearningFeature EngineeringPyTorchDeep LearningJuliaMachine Learning Operations
Deep Reinforcement Learning
Expertise in building deep reinforcement-learning agents using Soft Actor-Critic, Multi-Agent RL, Hierarchical RL, Reward Engineering, State/Action Design, Policy Evaluation, Constraint Handling, and Ablation Studies, with practical application to energy systems including demand response and energy storage.
Programming & Machine Learning
Proficiency in Python, JAX, Flax, PyTorch, NumPy, Pandas, Scikit-learn, GPU Vectorisation, Git, and SQL, enabling the development of scalable ML workflows, simulation pipelines, and data-driven energy models.
Energy Systems & Physical Modelling
Strong knowledge of power systems, energy storage, V2G/V2V, demand response, physics-grounded simulation, and optimisation, applied to sustainable energy communities, renewable integration, and smart infrastructure.
Time Series Analysis & Forecasting
Capabilities in sensor and smart-meter data processing, missing-data treatment, feature engineering, drift analysis, forecasting, conformal prediction, SHAP, and LIME for energy demand and renewable generation modelling.
Anomaly Detection & Predictive Maintenance
Experience developing machine-learning workflows for energy-asset monitoring, anomaly detection, demand forecasting, and operational-risk assessment, including asset allocation and sensitivity analysis.
Data Engineering & SQL
Hands-on experience integrating market, weather, generation, consumption, and asset-performance data using Python and SQL pipelines, as well as managing and validating technical data for electrical-grid and substation engineering workflows.
Model Interpretability
Application of SHAP attribution, native variable-selection weights, and LIME local explanations to cross-validate model interpretability and support evidence-based energy planning and decision-making.
Demand-Side Optimisation & Renewable Integration
Research and practical contributions to demand-side optimisation, renewable energy integration, electrification impacts, energy flexibility, and sustainable energy-community planning.
Data Visualisation & Reporting
Delivery of Power BI monitoring dashboards and translation of model outputs into actionable recommendations for technical, operational, and commercial teams.
Expertise in building deep reinforcement-learning agents using Soft Actor-Critic, Multi-Agent RL, Hierarchical RL, Reward Engineering, State/Action Design, Policy Evaluation, Constraint Handling, and Ablation Studies, with practical application to energy systems including demand response and energy storage.
Programming & Machine Learning
Proficiency in Python, JAX, Flax, PyTorch, NumPy, Pandas, Scikit-learn, GPU Vectorisation, Git, and SQL, enabling the development of scalable ML workflows, simulation pipelines, and data-driven energy models.
Energy Systems & Physical Modelling
Strong knowledge of power systems, energy storage, V2G/V2V, demand response, physics-grounded simulation, and optimisation, applied to sustainable energy communities, renewable integration, and smart infrastructure.
Time Series Analysis & Forecasting
Capabilities in sensor and smart-meter data processing, missing-data treatment, feature engineering, drift analysis, forecasting, conformal prediction, SHAP, and LIME for energy demand and renewable generation modelling.
Anomaly Detection & Predictive Maintenance
Experience developing machine-learning workflows for energy-asset monitoring, anomaly detection, demand forecasting, and operational-risk assessment, including asset allocation and sensitivity analysis.
Data Engineering & SQL
Hands-on experience integrating market, weather, generation, consumption, and asset-performance data using Python and SQL pipelines, as well as managing and validating technical data for electrical-grid and substation engineering workflows.
Model Interpretability
Application of SHAP attribution, native variable-selection weights, and LIME local explanations to cross-validate model interpretability and support evidence-based energy planning and decision-making.
Demand-Side Optimisation & Renewable Integration
Research and practical contributions to demand-side optimisation, renewable energy integration, electrification impacts, energy flexibility, and sustainable energy-community planning.
Data Visualisation & Reporting
Delivery of Power BI monitoring dashboards and translation of model outputs into actionable recommendations for technical, operational, and commercial teams.
Experience
Professional experience10+ years
Freelancer since2016
Languages
ArabicGoodEnglishFluentItalianGood
Project history
Developed a hierarchical multi-agent reinforcement-learning controller in JAX using discrete Soft Actor-Critic to coordinate 30 heterogeneous households, EV charging, V2G/V2V exchange, and shared energy storage. Formulated the control problem through state and action design, reward engineering, battery and mobility constraints, equipment efficiencies, and peer-energy settlement rules. Reduced annual community net cost by 29.2% against the strongest rule-based benchmark. Built a GPU-vectorised simulation and evaluation pipeline with optimisation and rule-based baselines, bootstrap confidence intervals, reward ablations, and reproducible multi-seed experiments. Designed a privacy-conscious decentralised architecture supporting future federated and edge deployment.
Conducting doctoral research in Sustainable Development and Climate Change, specialising in Technology and Territory. Research focuses on decarbonisation, energy efficiency, renewable integration, smart infrastructure, urban energy systems, and AI-enabled sustainable energy solutions. Built energy modelling and forecasting workflows using Python, statistical methods, and deep learning to analyse electricity demand, renewable generation, EV demand, weather variability, and energy-system performance. Applied data-driven and simulation-oriented methods to support feasibility-style analysis, scenario evaluation, savings estimation, and evidence-based energy planning.
Developed machine-learning workflows for energy-asset monitoring, anomaly detection, demand forecasting, and operational-risk assessment. Integrated market, weather, generation, consumption, and asset-performance data using Python and SQL pipelines. Built optimisation and sensitivity-analysis workflows to evaluate asset allocation, utilisation, availability, and maintenance decisions. Delivered Power BI monitoring dashboards and translated model outputs into recommendations for technical, operational, and commercial teams.