Browse publications on Google Scholar (top-right) ↗
Resume (English only)
Research Experience
[{'Position': 'Research Intern', 'Institution': 'Google Research, Zurich, Switzerland', 'Time': 'January-April 2025', 'Project': "Working on a project to optimize long-context inference, improving LLMs' efficiency in processing and understanding extended inputs."}, {'Position': 'Doctoral Research Assistant', 'Institution': 'EPFL, Lausanne, Switzerland', 'Time': '2019 - 2025', 'Project': 'Contributed to multiple research projects on multilingual LLMs, covering the entire training pipeline. Led the multilingual effort within the SwissAI initiative. Collaborated on projects with Google, Cohere, and HuggingFace. Supervised junior researchers and summer interns.'}, {'Position': 'Scientific Assistant', 'Institution': 'Machine Learning and Optimization Laboratory, EPFL, Lausanne, Switzerland', 'Time': 'Sept. 2018 - Sept. 2019', 'Project': 'Involved in the mlbench project, a benchmark framework for distributed machine learning.'}, {'Position': 'Research Intern', 'Institution': 'Data Analytics Laboratory, ETH, Zurich, Switzerland', 'Time': 'May - July 201y', 'Project': "As an intern in Thomas Hofmann's lab working under the supervision of Carsten Eickhoff, worked on a modular, patient-centric information retrieval system designed for precision oncology applications. The result of the project was a submission to the TREC 2017 Precision Medicine track."}, {'Position': 'Research Intern', 'Institution': 'Max Planck Institute for Software Systems, Kaiserslautern, Germany', 'Time': 'February - April 2017', 'Project': 'As an intern under the supervision of Manuel Gomez Rodriguez, worked on a project analyzing the dynamics of citation networks: quantifying the value of a set of published papers and modeling knowledge diffusion across a citation network.'}, {'Position': 'R&D Engineer', 'Institution': 'Center of Intelligent Vision & Image Processing, Shiraz University, Shiraz, Iran', 'Time': '2016 - 2017', 'Project': 'Worked on projects focused on object detection, facial expression analysis, and real-time face recognition and tracking.'}]
Background
Research interests include natural language processing (NLP) and machine learning, with a particular focus on improving the multilingual capabilities of large language models (LLMs), especially in low-resource settings. Works across the entire pipeline of training multilingual LLMs.