Invoice Master: thousands of supplier invoices, sorted by AI

The challenge

Kenny Hills Hospitality Group receives supplier invoices from across its operations — thousands of them, fragmented across formats and impossible to sort manually at any reasonable cost. The group's only workable option had been to assign a full team of six people to the task for one to two weeks, every cycle. Beyond the cost, manual sorting at that volume invites human error into financial data.

The solution

Edmund Ng Young Hoong, Head of Digital & IT, brought us the problem statement. Over a few structured discussions we drafted the PRD together, defining exactly what the tool needed to read, extract and validate. Three days later, Invoice Master was a working tool: an AI-powered system that classifies, sorts and processes the group's supplier invoices automatically.

The results

• 3 days from problem statement to working tool — including PRD workshops.
• 6-person, 1–2 week manual cycle eliminated. Thousands of invoices are now sorted "in no time at all," without constant supervision.
• Reduced human error in invoice sorting and the financial data that depends on it.
• Hardened through edge-case testing. After launch we tested extensively against edge cases to verify extraction accuracy, then added AI-driven insight features to future-proof the tool.

In their words

"i12ai team has been extremely proficient and structured when it comes to developing a problem solving tech solution for Kenny Hills Hospitality Group. We had an issue where all the invoices we received from suppliers are fragmented and are impossible to sort out manually unless we use a full team of 6 to work on it for 1-2 weeks. With the solution that they have built, which is called the Invoice Master, I am now able to sort out thousands of invoices in no time at all, without even me to constantly be on my toes, thus reducing human error when it comes to sorting out a bunch of invoices.

Their after service was also phenomenal as we tested a lot of edge cases to ensure the data we received from the invoice sortation system is accurate. To top it off, they even went above and beyond by adding additional features to give more insights and leveraging on AI to finetune and enhance the tool for future proofing."

Why this worked

This is the Discover, Design, Deploy, Optimise method compressed into its fastest form. The client owned the problem statement; we translated it into a PRD in days, not months; the build targeted one high-value workflow instead of a sprawling platform; and post-launch optimisation (edge-case testing, insight features) came after the tool was already earning its keep. Document processing automation is one of the highest-ROI starting points for AI transformation — see our AI Implementation service or the F&B industry page for how the same pattern applies elsewhere.