Finance and admin
AI for comparing supplier quotes
Collect the quotes for one purchase, whatever their format. Ask AI to pull the same fields from each one: unit price, quantity, delivery time, payment terms, validity, warranty and extra charges. Put them in one comparison table and ask AI to flag what is missing or unclear. Check the key numbers against the original quotes, then make the decision yourself.
Effort:
Low
Time saved:
Roughly 1–3 hours per purchase comparison; measure it in a pilot
The problem today
Quotes arrive as PDFs, emails and spreadsheets, each with different units, terms and hidden extras, so comparing them takes hours and mistakes slip through. Large manufacturers and retailers already use AI in purchasing to compare offers and spot unusual terms. For a Greek SME the affordable version is an AI tool that turns the quotes you already receive into one clean comparison table.
How it works, in plain words
Imagine an assistant who lays all quotes side by side, converts them to the same units and circles the small print.
What goes in: the quotes for one purchase in any format, plus your fixed comparison table and the real quantity you plan to buy.
What the AI does: it pulls the same fields from each quote, converts units, adds transport and extras to show a total cost for your quantity, and notes where each value came from.
What a person checks: the buyer checks the key numbers against the original quotes, asks suppliers about anything missing or unclear, and makes the choice.
What comes out: one comparison table with totals, gaps and unusual terms flagged, ready for the purchase decision or a follow-up email to suppliers.
It can misread or invent figures and knows nothing about a supplier's reliability beyond what you tell it.
How it could work for businesses like yours
Illustrative examples, not client case studies.
A 60-person metal furniture manufacturer in Sindos
The purchasing manager compares steel and powder-coating quotes that use different units, minimum orders and payment terms. The company would use one fixed table for every purchase and let AI fill it, with the page reference for each value. It would aim to cut comparison time per purchase, as an estimate to verify, and track how often hidden extras are caught before ordering.
A cosmetics e-shop in Xanthi
The owner is choosing a new supplier for jars, labels and shipping boxes, with quotes in Greek, English and Italian. AI would convert all quotes to cost per 1.000 units at the real order size and flag missing delivery times. The owner would check the totals against the originals and measure whether the chosen offer matches its quoted cost on the first invoice.
A 7-person event marketing agency in Piraeus
For each trade show, the agency collects quotes for stand building, printing and staffing under tight deadlines. It would build a reusable comparison table per event and ask AI to list gaps, such as missing insurance or setup costs. A realistic first target could be a clear comparison sent to the client within a day of the last quote.
How to do it, step by step
Define what you compare: unit price, minimum order, delivery time, transport costs, payment terms, validity date, warranty and penalties. Save it as a fixed table you reuse for every purchase.
Upload the quotes for one purchase to a business AI tool and ask it to fill the table, with a note of where each value came from, such as the page or line.
Ask it to convert everything to the same units and to show the total cost for your real quantity, including transport and extras. Check the calculations yourself, because AI can misread or invent numbers (see AI hallucination).
Ask AI to list what is missing or unclear in each quote and to draft short, polite questions for each supplier. Never paste one supplier's prices into a message to another.
Decide with the people who know the suppliers. Quality, reliability and the relationship matter as much as the lowest price.
Save the table for each purchase. After a few months you have a simple price history per supplier and material.
Tools to consider
ChatGPT Team or Enterprise: reads PDFs and spreadsheets and builds comparison tables quickly. Caveat: it can misread numbers in scanned quotes, so always check prices against the original.
Microsoft Excel with Copilot: builds the comparison and total-cost formulas in the spreadsheet you already use. Caveat: it works best when the data is already in a table, so PDFs may need a first extraction step.
Claude: handles several long quotes at once and explains differences in terms in plain language. Caveat: you still need to check every figure it reports.
Data protection and GDPR
Supplier quotes are mostly business data, but they include contact names and sometimes personal data of sole traders. The bigger issue is confidentiality: quotes often contain prices the supplier does not want shared. Use business AI plans with a data processing agreement and training switched off. Keep each supplier's prices out of messages to others, including AI-drafted emails. Record the final decision and the reasons behind it, which also helps if the purchase is part of a project with public or EU funding. Check with your DPO or legal advisor where needed.
Mistakes to avoid
Comparing unit prices without adding transport, extras and payment terms.
Trusting figures pulled from scanned PDFs without checking them.
Letting the cheapest line in the table make the decision instead of the full picture.