Research scientist & AI engineer · Lisbon, Portugal

Reinforcement learning,
planning &
decision-making.

I study convergence in reinforcement learning, hierarchical decision-making, and adaptive planning. As an AI/ML engineer, I work on search, recommendation, conversational commerce, and agent evaluation.

Portrait of Diogo S. Carvalho
NowResearch Scientist &
AI/ML Engineer
Rezolve AI
PhD

Awaiting defense
Instituto Superior Técnico

01 / Research & engineering

Studying how agents learn and plan; building systems for search, recommendation, and agent evaluation.

Research scientist

Reinforcement learning across representations, timescales, and levels of abstraction.

My research examines convergence with function approximation and non-stationary features, hierarchical reinforcement learning, temporal abstraction, and when an agent should plan, act, or reuse a skill.

Research programme and publications →
AI/ML engineer

Search, recommendation, conversational commerce, and evaluated agentic systems.

At Rezolve AI, I work on retrieval and recommendation models, conversational product experiences, and trace-based evaluation of agent state, coordination, and execution failures.

Professional experience →

Latest

News & activity

Full research record →

Personal

I read, play games on Nintendo Switch, train, cook, and play Sunday football with Curti Zona FC. I keep a few current notes and links on a separate page.

Personal notes →