Financial and Business Intelligence Analyst with hands-on experience at Airbus Defence and Space, owning the full reporting cycle for a multi-team data and analytics portfolio. Track record of building financial trackers, KPI monitoring frameworks, and executive-ready dashboards that drive operational and financial decisions.
Tools: Excel | SAP Analytics Cloud | Power BI | Python | Google Workspace | Jira
Open to roles in: Virginia | North Carolina | Remote
Built a five page executive BI dashboard in SAP Analytics Cloud analyzing 24 months of fictional business performance data for a global technology company across 5 regions and 4 product lines.
- Designed a structured data model in SAC Modeler with custom calculated measures for variance analysis, gross margin, net customer movement, and a RAG performance framework with defined revenue variance thresholds of greater than 3% Green, between -3% and 3% Amber, and below -3% Red.
- Built interactive dashboards with consistent filtering across all five pages, covering executive summary, revenue performance, cost performance, profitability analysis, and year over year trends.
- Solved SAC platform limitations including negative value color scaling in heatmaps, alphabetical month sorting, and a headcount aggregation error on scatter plots caused by residual chart configuration, all resolved through iterative model and design workarounds.
- Applied standard financial reporting conventions throughout, including cost variance directional logic where negative variance represents favorable underspend.
Predicted hotel booking cancellations using real-world data to support revenue optimisation decisions.
- Cleaned and merged multi-source booking data, handling bias by dropping nationality to ensure fairness.
- Analyzed correlations to identify the top five drivers of cancellations: lead time, prior cancellations, booking changes, parking requests, and special requests.
- Trained and compared four classification models: Logistic Regression, KNN, Decision Tree, and Random Forest.
- Achieved the best overall balance with Random Forest at 82% F1 score, while Decision Tree delivered highest recall at 78%, minimizing missed cancellations.
- Presented actionable insights for hotel revenue management and overbooking risk reduction.
Analyzed how modern language models interpret word meaning through vector-based text representations.
- Implemented Skip-Gram and CBOW architectures using Word2Vec to compare how each learns context from surrounding words.
- Trained embeddings on a text corpus to visualize relationships between semantically related words.
- Applied dimensionality reduction techniques including PCA and t-SNE to show how similar words cluster in vector spaces.
- Analyzed learned embeddings to explain how algorithms capture linguistic context for downstream NLP tasks such as sentiment analysis and topic detection.
- Gained deeper understanding of data representation, vector similarity, and the logic behind modern language models.
I build things that help people make sense of complex data. The tools change but the goal stays the same: turning complexity into clarity.