Yaroslav

Halkin

AI-assisted demonstration project using fictional order data. A Python script normalizes dates, prices, categories and statuses, removes exact duplicates after normalization, and records invalid or conflicting rows with rejection reasons. Example: 27 input rows, 18 accepted, 3 duplicates removed and 6 rows flagged for review. The original input is preserved. A separate Excel report summarizes accepted paid orders by month and category. Python produces CSV/JSON outputs; the Excel workbook does not automatically refresh from a new CSV. The project includes sample inputs, outputs, instructions and six data-quality tests. This is a portfolio demo, not paid client work.

AI-generated concept study using a fictional, unbranded water bottle. One presentation sheet compares three treatments: a clean catalog background, a natural lifestyle setting and a studio campaign scene. Generated from a text brief and reviewed for presentation and visual consistency. No customer photos were used. This is a portfolio demo, not paid client work. For real products, source photos and required details must be reviewed and product fidelity tested before confirming scope and delivery.