Data analyst in Edinburgh focused on SQL, decision-ready analytics and reliable data products. I have an MSc in Data Science from Heriot-Watt University and an engineering background in manufacturing.
My projects start with a business or operational decision, make the data grain and metric definitions explicit, and report limitations alongside results. I also build evaluated AI systems when AI is genuinely useful to the problem.
Airflow and dbt pipeline feeding a governed partner-performance layer. The project models seller-order and seller-month grains, reconciles revenue, tests service metrics and turns the analysis into an executive partner-management recommendation.
SQL dbt Airflow DuckDB Data quality Partner analytics
Country-year analysis across 202 countries with SQL, forecasting benchmarks, clustering and self-contained dashboards. The naive persistence forecast beat the machine-learning models, and the project explains why that is the useful result.
Python SQL Forecasting Dashboarding Statistical analysis
Fifty years of Bank of England decisions joined to ONS inflation data, with a SQL schema and interactive dashboard. The analysis distinguishes descriptive correlation from causal claims.
Python SQL Streamlit Time series Official data
Operational risk model trained on 96,000 marketplace orders using a chronological split, leakage-safe preprocessing and an action-sized review threshold. Served via FastAPI with tests, Docker and CI.
Python scikit-learn FastAPI Docker Model evaluation
- Materials-Informatics Copilot - deployed tool-using assistant with grounded quotations, 100 automated tests and a separate live evaluation.
- Retail NL2SQL Assistant - guarded natural-language analysis over a seven-table retail star schema with parsed SQL, read-only execution and auditable result rows.
Analytics: SQL, Python, pandas, cohort and funnel analysis, A/B testing,
forecasting, Power BI, Excel
Data platforms: dbt, Airflow, DuckDB, PostgreSQL, BigQuery, dimensional
modelling, data-quality testing
Engineering: FastAPI, Docker, pytest, GitHub Actions, Streamlit
Applied AI: RAG, structured extraction, tool-calling agents and LLM evaluation