A customer enquiry arrives by email; a sales quotation exists in SAP Business One minutes later — that is what this demo shows. TEKAI reads the enquiry, extracts the customer, items and quantities, matches them against SAP Business One master data, and drafts the quotation for a human to approve. The typing disappears; the judgement stays.
Enquiries arrive as prose. Part numbers come misspelled, quantities sit inside sentences, and three items are scattered across two paragraphs. Someone has to read the email, interpret what the customer actually wants, look up the item codes, check the customer record and key the quotation line by line. On a busy day that is hours of latency between “customer wants to buy” and “customer has a price” — and speed of quotation is routinely the difference between winning and losing SMB orders. The customer who receives a correct, complete quotation in minutes concludes you are organised before a single unit ships; the one who waits until tomorrow may already have ordered elsewhere. The frustrating part is that none of this latency comes from hard decisions. It comes from transcription — moving facts out of an email and into the fields of a document — and it consumes exactly the people who should be spending the day selling.
Running on a demonstration database: an enquiry email goes in, and TEKAI does the pass a good inside-sales person would. It identifies the business partner behind the email address, resolves the described items to item-master codes, reads the quantities out of the prose, applies the price list assigned to that customer, and produces a draft sales quotation in SAP Business One — numbered, complete, ready for the sales order that follows. The salesperson reviews the draft and sends it.
Just as important is what happens when the email is not clean. Ambiguities — an item description that matches two codes, a quantity that reads oddly — surface as questions for a person, not as silent guesses baked into a document. Every figure in the demo comes from a demonstration database, not customer data; the workflow is the same on your own enquiry inbox.
Email-parsing templates have been tried before, and they break on the third differently-worded enquiry — a new phrasing, a reordered list, a colloquial product name, and the rule set falls over. A language model grounded in your live item and customer masters handles that variation the way your best inside-sales person does: it reads meaning against real data, not patterns against a form. The grounding is the point. The model is not guessing what “the usual 20mm fittings” might be; it is resolving the phrase against the items this customer actually buys.
The AI drafts; authorised users approve. Nothing goes to the customer without a person deciding it should. Pricing overrides, credit checks and special terms remain exactly where your approval process puts them today — the AI removes the keying, not the control. That matters most on the edge cases: the customer near a credit limit, the enquiry that asks for a discount, the item with thin margin. Those are the moments a business earns or loses money on judgement, and the workflow is built so they still land on a person’s desk — with the draft already prepared, not a blank screen. The propose-and-approve pattern is the same one that runs through every TEKAI workflow.
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/).
Yes — document drafting like quotations is part of the workflow, with human approval before anything is sent.
The AI asks instead of guessing — unmatched items and odd quantities come back as questions for a person to resolve.
Yes — it applies the customer’s assigned price list from SAP Business One.
The same extraction-to-quotation flow applies to messages — the channel is an input, the workflow is identical.
In the demo, minutes from email to draft — review time is then yours. The fastest way to see it on your own data is the five-minute Fit Test.
New to the category? Start with the overview: AI for SAP Business One, explained.
Update, August 2026: TEKAI now also works with Gemini, Microsoft Copilot, and Grok.