I have a sizeable collection of banking data made up exclusively of loan transactions. I want to put this data to work by applying AI-driven analytics that uncover useful business insights. Your task is to take the raw transaction records, prepare them for analysis, and deliver clear, data-backed findings that I can act on. Scope • Ingest and clean the loan-transaction dataset (formats are CSV and SQL export). • Apply appropriate statistical or machine-learning techniques to surface patterns, anomalies, trends, and any other insights that add value. • Present the results in a concise written report and a reproducible notebook or script (Python preferred, but I am open to R or other industry-standard tools) so I can rerun the analysis with future data. • Include brief comments or markdown cells explaining each major step, from preprocessing through to interpretation of results. Deliverables 1. Fully commented code/notebook. 2. A short executive summary (PDF or DOCX) highlighting key findings. 3. Optional data visualisations or dashboards if they help communicate the insights more effectively. Acceptance Criteria • All code executes without errors on my sample dataset. • Findings are clearly tied back to the underlying data and illustrated with relevant charts or metrics. • Documentation is complete enough for a non-technical stakeholder to grasp the headline insights and for a technical user to replicate the workflow. If you have experience with banking datasets, model interpretability, or building lightweight dashboards in tools like Tableau or Power BI, please mention it—those skills will be a plus.