Requests for quotations frequently arrive by email. Customers describe their requirements in the message body, use their own product descriptions, or provide additional information in PDF, Word, or Excel attachments. Sales employees must read the request, identify the customer and contact, match the requested products, recognize missing information, and manually create a quotation with all its line items in Dynamics 365.

In this session, I will demonstrate an end-to-end process in which an incoming request for quotation is processed with artificial intelligence and results in the automatic creation of a complete quotation in Dynamics 365.

The process begins with a new email arriving in a central sales mailbox. Power Automate retrieves the message body and any available attachments. A language model analyzes the request and converts the identified information into a predefined data structure. This can include the sender, company, delivery address, currency, requested delivery date, products, quantities, and units of measure.

The extracted information is then matched against the master data in Dynamics 365. The process checks whether the company and contact already exist and maps the customer’s product descriptions to internal item numbers. Alternative spellings, incomplete descriptions, and potential typographical errors can also be considered during the matching process.

Once the data has been identified and matched, the remaining quotation information is added. Prices, currencies, units, and other commercial information are not freely determined by AI. They are retrieved from the available master data, pricing information, and clearly defined business rules in Dynamics 365.

When all required information is available, Power Automate automatically creates the quotation header and the corresponding quotation lines in Dynamics 365. The original email and its attachments can be linked to the transaction, preserving traceability between the source request and the resulting quotation. Finally, the responsible sales employee can be notified and the quotation made available for review or further processing.

In addition to the live demonstration, we will examine the underlying architecture involving Outlook, Power Automate, AI services, Dataverse, and Dynamics 365. Other topics include structured AI outputs, customer and product matching, missing-information handling, confidence values, exception management, and determining which decisions can safely be automated.

Attendees will experience a complete, practical process: a request for quotation arrives by email, and only moments later the corresponding quotation—including header and line data—is available in Dynamics 365.

The session uses anonymized or fictional emails, attachments, customer records, product information, prices, and quotation data. No confidential customer-project information will be presented.

In the demonstrated scenario, the quotation is created automatically in Dynamics 365. For production use, organizations must decide—based on their data quality, business rules, and risk assessment—whether every quotation may be created automatically or whether uncertain or incomplete results require human approval.

AI interprets the request and supports the mapping of the included information. Prices, discounts, taxes, units, and other binding commercial values are retrieved from Dynamics 365 or determined through clearly defined rules. They are not freely generated by the language model.

This is not a sales presentation. In addition to the benefits, the session will cover failure scenarios, model limitations, data protection, traceability, and operational costs. Depending on the technical conditions at the venue, the solution will be demonstrated live or through a prepared recording.

Recommended duration: 45–60 minutes.
Target audience: Dynamics 365 and Power Platform professionals, solution architects, developers, and professionals responsible for sales or internal sales operations.
Level: Intermediate.