From enquiry e-mail to order: AI automation of the quotation process

01Situation

Several thousand customer enquiries arrive by e-mail every year. Each was handled by hand: contacting suppliers, requesting prices and lead times, calculating, writing the quote and, once accepted, transferring data into ERP, CRM and accounting. Information was spread across several systems, so each case took considerable effort.

02Challenge

Around 25 suppliers, each with its own pricing rules, minimum quantities, currency logic and lead-time buffers, plus compliance evidence such as RoHS for every quote. Existing systems had to stay, and customer and pricing data could not leave the company network.

03Solution

Specialised AI agents read incoming enquiries, extract part numbers, quantities and date codes, query suppliers automatically and interpret their replies, including PDF attachments. A rule-based calculation engine applies supplier-specific rules, and a generator produces the quote including compliance details. Staff approve every quote before it is sent. Once accepted, order confirmation, ERP and CRM are updated automatically.

04Implementation

• —E-mail agent on the existing mail server: detect, structure and log enquiries
• —Supplier agent: automated price and lead-time requests, parsing of varied replies
• —Calculation engine with configurable rule sets per supplier
• —Quote generation with versions, compliance evidence and an approval workflow
• —Sync with ERP and CRM, data prepared for accounting
• —Internal language model on dedicated hardware, role-based access and full audit trail

05Impact

• —Significantly less manual work per quote
• —Faster responses to customers and suppliers
• —No more duplicate data entry across systems
• —Full data sovereignty: nothing sent to external cloud services