PhD student working on search agents

I build and study systems that search, read and reason over large collections of text. I care about how an agent reaches its answer, not just about the answer: how to measure that, and how to train models that hold up when the evidence is poor.

Keywords: Search Agents, Deep Research, Dynamic Environments

Orange Innovation · Lannion, France ·ismael.rousseau@orange.com

layer 01

Selected publications

All publications →

DeepStress: Stress-Testing Deep Search Agents

Ismaël Rousseau, Géraldine Damnati, Frédéric Béchet

WAB @ COLM 2026 · 2026

O_FT@EvalLLM2025 : étude comparative de choix de données et de stratégies d'apprentissage pour l'adaptation de modèles de langue à un domaine

Ismaël Rousseau, Claire Perroux, Pierre Adam, Thomas Girault, Lionel Delphin-Poulat, Morgan Veyret, Gwénolé Lecorvé, Géraldine Damnati

EvalLLM 2025 · 2025

Darbarer @ AutoMin2023: Transcription simplification for concise minute generation from multi-party conversations

Ismaël Rousseau, Loïc Fosse, Youness Dkhissi, Géraldine Damnati, Gwénolé Lecorvé

INLG 2023 — Generation Challenges · 2023

Exploring Social Sciences Archives with Explainable Document Linkage through Question Generation

Elie Antoine, Hyun Jung Kang, Ismaël Rousseau, Ghislaine Azémard, Frédéric Béchet, Géraldine Damnati

LaTeCH-CLfL 2023 @ EACL · 2023

layer 03