From new supplier to payment
Nine steps, five roles. The yellow lane is the purchasing associate. Each step shows what is stuck there this week. Select a step to see what happens and what the associate does.
Open orders that need someone to act
Every row has an owner and a next action. The filters work like report slicers: everything on this page follows them.
Action items by buyer
Select a buyer to filter
First three things today
Picked by impact
Invoice price against PO price
The 3-way match compares PO, goods receipt and invoice. Tolerance here is within 2% and no more than $50 for the invoice. Inside it, the system releases the invoice. Outside it, the invoice goes on hold until someone fixes the cause.
Why prices did not match
Last 12 weeks
What to do for each cause
Setups, supplier issues and payment requests
The follow-up side of the job: get new suppliers set up safely, answer suppliers with facts and a date, and send payment requests that do not bounce.
New supplier setups in progress
Bank details are confirmed by phone, on a number found independently of the email: an existing contact, the contract, or the main switchboard. Never a number from the same email. Bank-change emails are a common fraud attempt.
Supplier issue log
Communication failures, payment questions, late confirmations
Payment requisitions: check before submitting
Payments that do not go through a PO
Approval rule in this demo: up to $10,000 manager, above $10,000 director.
What the procedure says against what the data shows
Operating procedure targets
Black line = target · bar = actual
Where confirmations break
PO confirmed within 2 business days
Late or missing confirmations by supplier, top 8. email-only = not on the supplier network; POs go by email.
Why invoice holds miss 5 days
Cleared within 5 business days, by reason
Where supplier setup time goes
Average days per stage, completed setups
Findings and fixes
Ordered by impact
Show the SQL behind the confirmation analysis
Data, Power BI measures, and where AI helps
Data model
Same shape as standard purchasing exports
- Suppliers 1 → many PO lines, on
supplier_id. Supplier ID must be unique, or totals double. - PO lines 1 → 0 or 1 invoice, on
po_line(PO number + line). - Buyers 1 → many PO lines, on
buyer_id. A date table joins onpo_date.
How I'd rebuild it in Power BI
DAX measures
Calculated columns first, then measures
When the numbers don't match
- Same filters and date range as the number you're comparing to?
- Last refresh time. Did the source file move or get renamed?
- Relationships. A supplier with two remit-to rows makes the join many-to-many and doubles totals.
- Tie one supplier's total back to the source system by hand.
Where AI helps
AI drafts, a person checks and decides