Staff AI/ML Engineer · Ph.D. in Physics · Lisbon, Portugal 🇵🇹
I build ML systems that operate on large-scale, high-velocity data — currently at EY-Parthenon, shipping agentic systems, LLM integrations, and platform-level data pipelines in the Europe West AI Hub. Before that, I spent four years at EDP on time-series forecasting, portfolio optimisation, and LLM-powered automation for energy markets, and six years at CERN's CMS experiment hunting for supersymmetry and tinkering with quantum machine learning.
👉 edge-cavai.diogobastos.workers.dev — a RAG-powered chatbot that answers questions about my background. Ask it anything.
- Agentic systems & GenAI — agentic systems, LLM integrations, and platform-level pipelines that dozens of teams depend on; Azure OpenAI, Llama, LangChain, RAG, Weaviate; report-automation assistants that cut manual review time >40%
- Forecasting & optimisation — LSTM / CNN / Temporal Fusion Transformers for power-price prediction (cut error from 15% → 5%); quadratic programming for portfolio allocation (+30% gross margin)
- MLOps — Databricks, Python ecosystems, pipelines that survive contact with reality
Ph.D. thesis: Search for top squarks in the four-body decay mode with single-lepton final states at the LHC — published in JHEP06(2023)060. Side quest: Variational Quantum Classifiers and Quantum GANs (PennyLane, Qiskit) for Monte Carlo augmentation.
Python · PyTorch · TensorFlow · scikit-learn · LangChain · SQL · Databricks · Azure · C++ · Bash