Fabio Petroni
Rome, Italy
 

I am a Group Leader at the European Molecular Biology Laboratory (EMBL) in Rome since September 2025, focusing on AI-driven scientific discovery.

I am affiliated with the European Bioinformatics Institute (EMBL-EBI) since March 2026. 

I am also Co-Founder at Samaya AI from 2022, building an AI-powered knowledge-discovery platform.

From 2018 to 2022, I was a Researcher at the FAIR (Fundamental AI Research) team of Meta AI (formerly known as Facebook AI Research), where I focused on Knowledge-Intensive Natural Language Processing.

Before that, I was a Research Scientist at the R&D department at Thomson Reuters and a Data Engineer at KMPG.

I obtained my PhD from Sapienza University where I was part of the Research Center on Cyber Intelligence and Information Security. My thesis advisor was Prof. Leonardo Querzoni.

During my PhD I've been a Visiting Researcher at the University of Mannheim in the Data and Web Science Group under the supervision of Prof. Rainer Gemulla.

 

Invited talks:

- Panelist at ROMA.AI (September 2025), on Generative AI: “Allucinazioni, Legge di scala, Small Language models: limiti ed opportunità per l’AI generativa e l’Agentic AI”

- Keynote talk at Eval4RAG: Workshop on Evaluation of RAG Systems at ECIR 2025 (April 2025) on "A Journey Through Domain-Specific RAG and Agent Evaluation"

- Invited talk at Sapienza DIAG Seminar Series (October 2024), title: “Retrieval Augmented Generation (RAG): Applications, Limitations, and Future Directions”

- Invited keynote at the KBC-LM workshop @ ISWC (November 2023), title: “Can machines assist humans in discovering new knowledge?"

- Panelist at the Italian Tech Week 2023 (September 2023) on a “new generation of Italian entrepreneurs”

- Panelist at the Workshop on Generative Information Retrieval @SIGIR23 (July 2023) on “Model Behavior”

- Invited talk at the London Information Retrieval Meetup (June 2023), title: “Autoregressive Knowledge Retrieval”

- Invited talk at the Edinburgh ILCC Seminar Series (April 2023), title: “Autoregressive Knowledge Retrieval”

- Invited talk at London AI (November 2022), title: “Improve Wikipedia Verifiability with AI”

- Invited talk at NL4AI (November 2022), title: “Improve Wikipedia Verifiability with AI”

- Invited talk at DeepMind (November 2022), title: “Autoregressive Knowledge Retrieval”

- digitaltrends interview (August 2022)

- PointerPodcast interview (in Italian, August 2022)

- Il Messaggero interview (in Italian, July 2022)

- Invited talk at UCL AI Society Journal Club (18 January 2022), title: “Verifying Citations with Context-Aware Natural Language Processing”

- Invited talk at AI Ukraine (30 October 2021), title: “What do machines know?”

- Deep Learning and Natural Language Processing lesson for the Data Mining Technology for Business and Society course of Prof. Stefano Leonardi at Sapienza, University of Rome (15 May 2021)

- Invited talk at JEEC|21 (19 March 2021), title: “What do machines know?”

- Invited talk at Stanford MLSys Seminar Series (11 March 2021), title: “Assessing Machine Knowledge”

- Panelist at the Knowledge Connexions Conference on AI + Knowledge - a match made in heaven? (30 November 2020)

- Invited talk at Unstructured and Structured KBs workshop (25 June 2020), title: “How can we compare unstructured, structured and self-structured knowledge representation?”

- Deep Learning and Natural Language Processing lesson for the Data Mining Technology for Business and Society course of Prof. Stefano Leonardi at Sapienza, University of Rome (14 May 2020)

 

Awards:

- Best Poster in the NeurIPS 2021 ENLSP Workshop for "Cutting Down on Prompts and Parameters: Simple Few-Shot Learning with Language Models"

- 1st place at the NeurIPS Efficient Open-Domain Question Answering competition 2020 for the 6GB constrained track and unconstrained track

Google Best Paper Award at AKBC 2020  for "How Context Affects Language Models' Factual Predictions"

Best Paper Award Runner Up at DIMWA 2019 for "SAFE: Self-attentive function embeddings for binary similarity"

- Research grant for the "Avvio alla Ricerca" (research launch) project from Sapienza (2015), title "SARE: Scalable Algorithms for Relation Extraction with contextual data"

- Finalist at the Telecom Italia Big Data Challenge 2014. A. Anagnostopoulos, F. Petroni, M. Sorella: “COLITA: Collaborative Interest-driven Targeted Advertising”

- Enrolled in the "percorso d'eccellenza" (excellence path program) for MSc. in Computer Engineering: special program for top 10th students in the MSc, a.y. 2011/2012

- Award "laureato eccellente Sapienza" (excellent graduate Sapienza), a.y. 2011/2012

 

DOWNLOAD MY FULL CV

 

HORTONWORKS HDP CERTIFIED APACHE SPARK DEVELOPER