// Data Scientist & AI Developer
Pooya
Sabbagh
Building intelligent systems at the intersection of economics, machine learning, and agentic AI. Based in Milan, Italy.
Scroll
// Data Scientist & AI Developer
Building intelligent systems at the intersection of economics, machine learning, and agentic AI. Based in Milan, Italy.
About
I am an MSc student in Data Science for Economics at the University of Milan, with an Erasmus semester at the University of Geneva. My background bridges rigorous quantitative analysis with hands-on AI engineering.
I build end-to-end AI systems — from autonomous agents with live web access to NLP retrieval pipelines — with a practical understanding of the business and policy context behind the data.
I am passionate about applying these tools to meaningful problems: food safety, environmental economics, and evidence-based policy are areas that genuinely motivate me.
Download CVProjects
An end-to-end agentic AI system built on a locally hosted 20B-parameter LLM with a 70,000-token context window. Uses the Model Context Protocol (MCP) to drive live Playwright browser automation — executing multi-step reasoning workflows for dynamic web data extraction and structured intelligence synthesis. Achieved sub-15-second automated reasoning cycles with zero API cost through GPU-offloaded local inference.
End-to-end text retrieval system using dense vector embeddings (SentenceTransformers) and hybrid scoring over a normalised SQLite database of 2,457 books. Achieves sub-second latency against natural-language queries via cosine similarity on precomputed embeddings.
Led a 9-person multicultural team to analyse brand positioning in the energy market. Built an automated Python/Selenium pipeline extracting and structuring consumer sentiment across 1,274 Trustpilot pages, navigating dynamic content and anti-bot mechanisms.
Designed a stochastic game environment formulated as a Markov Decision Process. Applied the Value Iteration algorithm to compute a deterministic optimal policy maximising expected cumulative rewards over a probabilistic state space.
Skills
Combining a strong quantitative foundation with applied AI engineering. I work across the full stack — from data acquisition and processing to model deployment and communication of results.
Education
Contact
Whether you're looking for a data science collaborator, a research assistant, or want to discuss a project — I'd be glad to hear from you.
Currently available for internships, research assistant positions, and freelance projects in data science, AI, and analytics. Based in Milan — open to hybrid and remote arrangements across Italy and Europe.
Send a message