Same item, different vendors, prices up to 100% apart — sitting unnoticed in purchase history. This demo shows TEKAI sweeping live SAP Business One vendor data and surfacing the gaps: which items you are buying at wildly different rates, from whom, and what the spread costs.
Purchasing sees one PO at a time. Each individual order looks reasonable — a plausible vendor, a plausible rate, an approval signature. The gap only exists across POs: this month’s rate versus last quarter’s, vendor A versus vendor B, plant 1 versus plant 2. Nobody runs that comparison manually across a full item master, because it means lining up every purchase of every item over a year and reading the spread — days of spreadsheet work with no guarantee of finding anything. So the comparison never runs, the spread persists, and it is pure margin leaking: money paid above the rate your own history proves another vendor accepts. Rate agreements drift, one-off urgent buys quietly become the new normal, and a vendor who quoted sharply to win the account edges prices upward once the ordering becomes routine. None of it is visible one PO at a time.
On a demonstration database: one conversation sweeps the purchase history and returns items ranked by price variance — the same SKU bought at rates up to 100% apart — with vendor-wise and period-wise breakdowns available on follow-up. Which vendor charged what, when, at which quantities; how the rate moved over the year; where the outliers cluster. The output is not a report to be studied — it is a negotiation list, ordered by money. The 100% figure is what the demonstration dataset surfaced; every number in the video comes from a demonstration database, not customer data. Your own spread is whatever your live purchase history shows.
A ranked list of gaps is only worth what you do with it, and each gap is an action. Consolidate volume to the better vendor. Renegotiate with the outlier, history in hand — “you charged us X in March and Y in August for the same item” is a strong opening position, and the follow-ups in the demo run the way a purchase review actually does — take one item, lay out its purchase lines, and the negotiation brief writes itself from your own documents. Or understand the legitimate reason for the difference — lot size, credit terms, urgency premiums — and price it consciously instead of paying it by accident. The AI proposes the list; purchasing makes the calls. The judgement about which supplier relationships matter and which premiums are worth paying stays exactly where it belongs.
The sweep that took this demo one conversation can run monthly. New gaps then surface while they are one PO old, not one year old — a fresh outlier rate gets questioned before the second and third orders land at the same inflated price. Price discipline stops being an annual project and becomes a standing check, and purchasing time goes where it pays: into the calls, not the collation.
SAP Business One (SQL or HANA) · TEKAI at ₹5,000/company/month India, $150 global · your own AI subscription (Claude, ChatGPT or Perplexity) · permission-scoped access, nothing trained on your data (/tekai/security/).
It surfaces the spread with the context — lot sizes, dates, terms — so purchasing judges legitimacy with the facts visible.
Yes, within whatever scope permissions allow.
No — the demo runs on a demonstration database; your own spread is what your live data shows.
Connect and ask for your top 20 items by purchase value with price variance — one conversation. The five-minute Fit Test tells you first whether TEKAI fits your setup.
The same check can run on schedule as an exception alert.
New to the category? Start with the overview: AI for SAP Business One, explained.