Looks like you have JavaScript disabled. For the full Twine experience, you will need to re-enable it.

Hi, I'm Alessandro Romeo, a passionate computer engineer specializing in artificial intelligence and data analytics. I enjoy designing innovative solutions and have hands-on experience developing semantic drift detection methods to improve language models. I'm a sociable and empathetic person who thrives in multicultural environments and teamwork. With solid academic training and practical experience in software engineering, machine learning, and system programming, I love building meaningful tech products. I'm always eager to listen, learn, and collaborate with diverse teams to tackle challenging problems creatively and efficiently.…Hi, I'm Alessandro Romeo, a passionate computer engineer specializing in artificial intelligence and data analytics. I enjoy designing innovative solutions and have hands-on experience developing semantic drift detection methods to improve language models. I'm a sociable and empathetic person who thrives in multicultural environments and teamwork. With solid academic training and practical experience in software engineering, machine learning, and system programming, I love building meaningful tech products. I'm always eager to listen, learn, and collaborate with diverse teams to tackle challenging problems creatively and efficiently.

Alessandro Romeo





Hi, I'm Alessandro Romeo, a passionate computer engineer specializing in artificial intelligence and data analytics. I enjoy designing innovative solutions and have hands-on experience developing semantic drift detection methods to improve language models. I'm a sociable and empathetic person who thrives in multicultural environments and teamwork. With solid academic training and practical experience in software engineering, machine learning, and system programming, I love building meaningful tech products. I'm always eager to listen, learn, and collaborate with diverse teams to tackle challenging problems creatively and efficiently.…Hi, I'm Alessandro Romeo, a passionate computer engineer specializing in artificial intelligence and data analytics. I enjoy designing innovative solutions and have hands-on experience developing semantic drift detection methods to improve language models. I'm a sociable and empathetic person who thrives in multicultural environments and teamwork. With solid academic training and practical experience in software engineering, machine learning, and system programming, I love building meaningful tech products. I'm always eager to listen, learn, and collaborate with diverse teams to tackle challenging problems creatively and efficiently.

Available to hire

Hi, I’m Alessandro Romeo, a passionate computer engineer specializing in artificial intelligence and data analytics. I enjoy designing innovative solutions and have hands-on experience developing semantic drift detection methods to improve language models. I’m a sociable and empathetic person who thrives in multicultural environments and teamwork.

With solid academic training and practical experience in software engineering, machine learning, and system programming, I love building meaningful tech products. I’m always eager to listen, learn, and collaborate with diverse teams to tackle challenging problems creatively and efficiently.

Skills

Experience Level

Expert

Expert

Expert

Intermediate

Intermediate

Intermediate

Intermediate

Language

Italian

Fluent

English

Advanced

Work Experience

Student Collaborator based on merit at Polytechnic of Turin

July 25, 2023 - July 21, 2025

Performed website updating and upgrading for the university, gathered and managed data and information, fostered multicultural relations, and worked collaboratively in a team-oriented environment.

Thesis student at Polytechnic of Turin

April 1, 2025 - Present

Designed and developed Semantic Drift Detection (SDD), a zero-shot method identifying semantically inconsistent or hallucinated outputs from large language models like GPT-3.5 and Mistral. Built a full experimental pipeline using inverse question generation, semantic similarity (SBERT), and a supervised Random Forest classifier. Generated and expanded QA datasets with automatic labeling and entailment-based corrections. Benchmarked the method against state-of-the-art approaches on datasets like TruthfulQA. Evaluated results using explainability, accuracy, precision, recall, and AUC, with automated performance reporting and visualization.