Internal Tool
AI-powered ETL tool for restaurant sales intelligence
An internal ETL tool built for Caffe Pasticceria that connects to their Point of Sale system via API, extracts transaction data, and transforms it into actionable business reports. The tool uses AI to generate human-readable summaries — giving the management team clear visibility into best-selling items, revenue trends, and peak hours without needing technical expertise.
The restaurant's POS system stored transaction data but offered limited reporting capabilities. Management lacked visibility into their own sales patterns — which dishes sell best, what times drive the most revenue, how categories perform over time. They needed actionable insights presented in plain language, not raw spreadsheets.
We built an ETL pipeline that connects to the POS API and extracts transaction data on a scheduled basis. The data is cleaned, normalized, and aggregated into structured datasets. AI processes the raw metrics and generates plain-language summaries that read like a business briefing — highlighting trends, anomalies, and opportunities. The reports are delivered through a clean dashboard the management team can access without any technical knowledge.
The ETL pipeline connects to the POS system's API for automated, scheduled data extraction. Data transformation layers clean and aggregate raw transaction records into structured datasets with category breakdowns, time-series trends, and item-level analytics. AI-powered report generation converts raw metrics into human-readable summaries using natural language processing. The dashboard presents visualizations, trend charts, and exportable reports.
The tool now provides the management team with regular, AI-summarized sales reports — giving them visibility into their business data that their POS system alone couldn't offer. Decision-making shifted from gut feeling to data-driven insights.