From Two Hours to Minutes: Automating Tourism Quotations

The Problem Before

Every quotation cost roughly two hours of senior staff time: manually cross-referencing hotel rate sheets, guide fees, entrance ticket prices, and a live exchange rate, then hand-assembling a Word document. A single quote touched upward of 50 data points across disconnected spreadsheets, and pricing in EUR while operating in TRY meant the exchange rate could go stale before the document was even finished.

What MPS Built

A Supabase-hosted PostgreSQL database made every supplier rate queryable from one authoritative source. A dedicated n8n workflow fetches and snapshots the live EUR/TRY rate at generation time. A rules engine applies margins and assembles pricing automatically, and a Python/python-docx step fills 56 token placeholders in a branded Word template, all coordinated across twelve-plus automated workflows.

Decisions That Matter

Quotation records are immutable by design: a PostgreSQL trigger freezes the rate snapshot, supplier selections, and final price at generation time, creating an audit trail for free. Multi-language output (Spanish, English, Turkish) is built into the data model, and sequential MPS-YYYY-NNNN reference numbers make every document traceable.

The Result

The quotation process now runs roughly 24× faster, and frees about 20 hours of operations time per week. The error vectors from manual rate lookups and copy-paste are gone, the exchange rate is always current and recorded, and the pipeline handles parallel quotation runs without degradation.

• n8n
• Supabase
• PostgreSQL
• Python
• python-docx