Research

What I work on

Four lines of work, each with what has been published on it. The full list is under Publications, and citations are on Google Scholar.

2026

Agentic AI: autonomy earned on evidence

When software may act on its own, and what it must show first: calibrated decisions, causal evidence, off-policy evaluation, guardrails inside the action, verification, and governance treated as engineering.

2024–2026

Decision systems for retail

Agents for pricing, inventory, assortment and allocation, from the retail agent operating model to production. Much of it comes from the data and AI teams at ALDO Group, and it continues at OODARIS AI.

2015

Software measurement and effort estimation

Choosing a software effort estimation model by its accuracy and its uncertainty, from the doctoral work at École de technologie supérieure.

2012–2015

Mobile application usability

How to evaluate the usability of mobile apps and games, an expert-based framework for iOS apps, and predicting an app’s user rating from its usability evaluation.