One queue, described so it can be measured.
Everyone can read a quote. Nobody has time to line three of them up.
A purchase request lands with the supply chain team. Someone writes the request for quotation, sends it to the usual suppliers, and waits. The replies come back in different formats, different units and different currencies, and someone types them into a spreadsheet to compare them.
Most of the comparisons are routine. The week is spent chasing the suppliers who have not replied, and checking whether a price that looks good is worse than the one paid last quarter. By the time anyone asks why a supplier was chosen, the spreadsheet has been overwritten twice.
What it usually looks like
You are not short of people who can read a quote. You are short of time to line them up properly before someone has to decide.
One AI employee, hired on a job description the procurement lead writes.
Nothing new is installed for the team. The AI employee is connected to the procurement mailbox, the purchase requests and the contracts and price lists the team already uses, and only to those.
On day one it only reads and prepares. Every request it drafts and every comparison it builds waits for a person. Before anything runs, the procurement lead sets two things, and both are recorded as decisions with their name on them.
Requests for quotation go only to approved suppliers, and every message leaving the organisation is sent by a person, at every trust level.
It never places an order, accepts a supplier’s terms or adds a supplier. The buyer decides, and the decision carries the buyer’s name.
The buying system stays the system of record. Xalios prepares the request, the comparison and the reasoning. It does not commit any spend.
“Prepare requests for quotation from approved purchase requests, for a buyer to send to the approved suppliers. Read the replies, line them up against the contract and the last price paid, and prepare a comparison with a recommendation. Report to the procurement lead.”
Four steps. One of them a buyer’s, and it takes ten minutes.
One purchase request, from the moment it arrives to the moment the order is placed. Nobody spends time on it until there is something to decide, and what reaches the buyer is a comparison, not a pile of replies.
A plant raises a request for 2,000 steel brackets, needed in six weeks.
Drafts the request from the purchase request and the drawing, lists the three approved suppliers for this part, and holds it for the buyer to send.
Three replies, in two currencies and two units of measure. Converts them, checks each against the contract and the last price paid, and flags that the lowest quote misses the date by two weeks.
The buyer opens the comparison, sees the recommendation and the reason, and picks the second-lowest price because it meets the date. The buyer places the order, and the comparison is filed with all three quotes attached.
What lands with the buyer is a comparison and a reason. Not three PDFs and a spreadsheet.
What the auditor would read, and what the procurement lead reads on a Friday.
The same record, with no separate reporting layer. Both can be replayed. Six months later the quotes, the contract rate and the price last paid open as they were when the choice was made, not as they are today.
For the supply chain team, twelve weeks in.
Not fewer people. Typing stops getting in the way of buying. Negotiation, new suppliers and the difficult calls stay with the buyers who are good at them, and the comparison is ready before the buyer opens the request.
Every request for quotation is written by hand, and every reply typed into a spreadsheet.
Requests are drafted from the purchase request, and replies are lined up as they arrive.
Quotes are compared on price alone, because the rest takes too long.
Price, lead time and terms side by side, checked against the contract and the last price paid.
Suppliers who do not reply are chased when someone remembers.
Follow-ups drafted on the date asked, and sent by a person.
Why a supplier was chosen lives in a spreadsheet that has been overwritten.
Every comparison carries its own record: the quotes, the reason and the buyer who decided.
Trust in automation is all or nothing, so it stays at nothing.
Authority is earned for each kind of purchase, raised by the procurement lead and withdrawn automatically when quality slips.
What stays with the supply chain team
The spend limit and the approved list
Both set and changed by the procurement lead. Every change is itself a record with a name on it.
The decision to buy
Which supplier, at what price, on what terms. The buyer decides, negotiates and places the order, and the decision carries the buyer’s name.
The buying system
It stays the system of record. Xalios prepares requests and comparisons with the reasoning attached, and does not commit any spend.
Open a comparison from week three and get week three.
Most questions about buying are questions about choice: why this supplier, at what price, and what else was on the table. A useful test at week twelve is simple. Pick ten purchases at random and ask for the comparison.
The answer to “why this supplier” becomes a click, not a search through old spreadsheets.
One AI employee becomes an operating model.
The memory built on quotations, which suppliers reply, which prices held, which parts are bought together, is the foundation the next role starts from. Each addition is faster than the last, and each one starts again at level 1 for its own kind of work.
Quotation desk
The pilot. Level 3 as the target, with chasing the silent suppliers as the next step you can choose.
Order confirmations
Checks each order acknowledgement against what was ordered and flags the differences before the goods ship.
Delivery exceptions
Watches for late and failed deliveries, gathers the proof and drafts the response for a person to send.
