Automated Purchasing

Automated Purchasing

Problem

At Poço, purchasing runs at scale: ~10,000 supplier price requests, 6,000–7,000 purchase orders and €12–13M in purchases every year.

• Supplier choice relies on people's know-how, not on the purchase-history data that already exists.
• ~7 days pass between an internal request and the purchase order.
• Delivery dates purchasing gets run 5 days later than the dates requested internally, so projects get delayed.
• Reading PDF quotes and comparing them line by line is slow, so decisions are made from too few quotes.

Solution

Automated Purchasing takes a purchase requisition all the way to purchase orders in one guided flow: Requisition → RFQs → Quotes → Comparison → Purchase orders.

How it works

1. Supplier pre-selection from purchase history. For each requisition line the usual suppliers are suggested. Suppliers can be added by hand for items with no history or no article in the database.
2. RFQs created and sent in bulk.
3. Supplier PDF quotes read in the browser. Each PDF is matched to its RFQ, and the app extracts grade, length, quantity/unit, price (converted to the requisition's unit), lead time and terms. Fields that need checking are flagged, and every value can be edited.
4. Side-by-side price comparison. The best price shows in green and real material alternatives are highlighted. Lead times, terms, validity and supplier delay history sit alongside.
5. Award per line, with partial ordering. You can order some lines now and leave the rest pending.
6. POs with proposed delivery dates. The date comes from the usual lead time in purchase history, or from the quote's lead time if that is longer. It is then moved to a weekday the supplier usually delivers on, and the buyer validates it.
7. POs created and sent in bulk.
8. Status at a glance. The requisition is colour-coded: green = ordered, orange = quantity differs, blue = RFQ pending. Lists of RFQs and POs are included.

Impact

Just 1% better purchasing through better negotiation or choosing from more quotes = €100k+ per year.

• More price requests: suppliers pre-selected from history, RFQs sent in bulk.
• Easier comparison: quotes read automatically and shown side by side.
• Shorter lead time to order: award per line, POs created and sent in bulk.
• Less delay vs requested dates and better project execution.