CV

Principal Data Engineer, Century Tech

• Operate in a "startup within a startup" environment, rapidly prototyping and iterating on agentic AI features.
• Build production-grade AI systems for automated feedback and assessment using Pydantic AI, context engineering, structured outputs and rigorous evaluation loops.
• Architect and ship full-stack products using FastAPI, Next.js, Railway and Neon, with end-to-end observability through Logfire.
• Establish spec-driven development practices with OpenSpec, aligning product design, implementation and AI-assisted coding workflows.
• Own the full product lifecycle, from problem definition and system design through deployment and scaling.

Senior Data Engineer, Century Tech

• Designed and built the company's data platform, including its data lake and Snowflake warehouse, from scratch.
• Built and managed Kubernetes-based infrastructure on AWS using Flux, Helm and Terraform.
• Developed and orchestrated core data pipelines using dbt and Dagster.
• Productionised machine-learning models by developing the MLOps workflows and infrastructure required to deploy and operate them.

Data Engineer / Senior Data Engineer, Kobalt Music

• Contributed to the migration from a legacy Oracle database to a modern data stack.
• Built ETL pipelines using Spark, Scala and EMR.

Big Data Engineer, IQVIA

• Built highly performant and scalable data systems on IQVIA's Big Data Platform (Spark, Hive, Impala, Hadoop).

Big Data Engineer, Cloudwick Technologies UK

• Data processing and analysis using Scala and Spark on AWS.
• Linux and Hadoop ecosystem administration (Hive, HBase, Sqoop, Oozie).

PhD in Computational Chemistry, University of Ferrara

• Developed new computational methods and scientific software (Python, Fortran).
• HPC programming and Linux systems administration.

Built an LLM-powered system providing instant feedback on student writing. Designed context engineering strategies and structured outputs to keep feedback consistent and useful. Iterated on evaluation and feedback quality based on real user behaviour. Deployed and maintained a production system used by real users.

Developed a fully agentic system for automated grading of multi-step maths problems based on mark schemes and reasoning processes. Focused on interpretability and alignment with human marking logic.

IoT platform for air quality monitoring, built in collaboration with the Italian National Research Council (CNR). Built the entire system and infrastructure: IoT sensors sending MQTT messages to TimescaleDB/PostgreSQL, backend and frontend for sensor management and data visualisation, and cloud infrastructure. particularmatter.org