About me

I’m seeking opportunities as an Applied Scientist, Research Scientist, or Machine Learning Engineer, specializing in NLP and IR.

I work at the intersection of natural language processing (NLP) and information retrieval (IR), focusing on areas such as conversational search, task-oriented dialog systems, document retrieval, and generative AI. I leverage state-of-the-art methods like retrieval-augmented generation (RAG), dense retrieval, performance-efficient fine-tuning (PEFT), and in-context learning with large language models (LLMs). I excel in collaborative environments and lead projects from ideation through deployment.

I hold a Ph.D. in Computer Science from USI, IDSIA in Lugano, Switzerland, where I conducted research on conversational AI under the supervision of Prof. Fabio Crestani. My research was centered on mixed-initiative conversational search. Specifically, I focused on methods for generating clarifying questions with an aim to better understand users’ information needs, as well as on LLM-based user simulation for improved conversational search evaluation.

I published in mulple top-tier NLP and IR venues, including SIGIR, ACL, WSDM, TIST, ECIR, COLING (700+ citations, h-index 10). For a full list of my publications, please visit my Google Scholar page. I actively contribute to the community as a program committee member for leading conferences such as SIGIR, WSDM, ECIR, CIKM, and EACL.

Research Interests

I’m interested in several research areas, related to NLP, IR, and LLMs:

• Evaluation of dialog systems – Exploring the complexity of user goals, understanding conversational context, and managing the diversity of possible conversational paths to achieve a goal.
• LLM-centric user simulation and sythetic data generation – Investigating these approaches as solutions for improving dialog system evaluation.
• Information retrieval (dense, hybrid) – Addressing challenges in retrieving long texts, product information, and multi-modal data.
• Answering complex information needs – Advancing conversational search and retrieval-augmented generation to tackle intricate user queries.

Machine Learning Engineering Interests

As a research engineer, I’m interested in deploying ML- and LLM-based solutions to real-world problems. These include:

• Efficient and scalable production-ready LLMs – Developing performance-efficient, task-specific fine-tuning methods and enabling the efficient deployment of LLMs in local or cloud environments.
• Agent-centric LLM-based task-oriented dialog systems – Designing policies and decision-making mechanisms using LLM agents.
• Dense retrieval beyond simple vector embeddings – Creating data-specific embeddings by fine-tuning appropriate models for better retrieval performance.
• Tackling tasks with traditional ML pipelines – Applying cost-effective traditional ML techniques for tasks that don’t require LLMs.
• Data analysis and cleanup – Emphasizing the importance of clean, well-understood data to avoid the “garbage in, garbage out” problem.

Feel free to reach out if you share similar interests and have ideas for collaborations!