About

I am Cesare Scalia, PhD: Lead Data Scientist at Life360, where I am responsible for causal measurement across advertising campaigns reaching millions of geolocated users. My work combines machine learning, causal inference, NLP and graph analytics with the design of AI-automation systems that compress analytical cycle time across an entire function.

Career arc

My first scientific life was in astrophysics. I completed a PhD in Physics at the University of Catania on stellar spectropolarimetry, with a visiting period at Armagh Observatory (UK), and continued briefly as an astronomer at INAF on the Cherenkov Astri-Horn telescope prototype. That decade gave me observation, instrumentation, data-reduction, parallel numerical computing, statistical-analysis and inversion/ML skills applied to stellar magnetic fields.

From research I moved into semiconductors, as a Device Engineer at STMicroelectronics, working on yield analytics across semiconductor manufacturing processes. The role grounded my statistics in a production environment with billions of measurements and unforgiving physics.

A short stint at Leonardo as a Data Science Intern applied PCA to stochastic systems and real-time object detection, bridging defence and industrial computer vision.

For five years at Eni I led a small team building AI-driven algorithms for corporate reputation analysis and measurement. The work integrated generative AI, NLP and Transformer models across computing, communications and public-relations sectors; I managed cloud development, ETL and AI-integrated reporting (AWS, SageMaker), and coordinated with external suppliers. In parallel I completed a postgraduate master in Economic Intelligence at IASSP, which re-framed my practice around strategy, geopolitics and open-source intelligence.

Today I lead data science at Life360 (ad-tech), a consumer platform serving advertising across millions of geolocated users. Since 2023 I have also consulted part-time for fantix.ai on Graph Machine Learning, multimodal link prediction, digital-twin synthetic data and look-alike models at scale.

What I build today

• Causal measurement. Uplift/CATE models, geo-experiments and synthetic-control incrementality designs for multi-million-user ad campaigns. The goal is always a defensible causal claim, not just a correlational dashboard.
• Agentic workflows. Custom Claude agents with tool use and MCP servers (Databricks, Atlassian) embedded in daily analytics, data-quality and pipeline-triage operations. These are not demos; they are the first responders for the analytics function.
• AI enablement. I author reusable Claude skills and Claude Code workflows, and lead firm-wide AI adoption across analytics teams through internal standards, onboarding playbooks and office hours.

Background

PhD in Physics (University of Catania, 2014–2017), visiting PhD at Armagh Observatory (2015–2016), with MSc and BSc in Physics from Catania (both 110/110 with honours). Postgraduate master in Economic Intelligence from IASSP (2022–2023). Peer-reviewed publications in A&A, MNRAS, Astronomical Journal and Proceedings of the IAU, with invited and poster contributions at IAU Symposium 340 and Machine Learning Prague.

Italian (native), English (C2), French (A2). Comfortable translating quantitative analysis into decisions for executive and board audiences.

Off-duty

Scuba diving, drum solos, open-source work, astrophysics (still) and time spent outdoors. Active in Rotaract, AVIS and Gruppo Astrofili Catanesi.