Problem Statement

Design, architect, and implement a widget that, at a high level, utilizes the latest AI tools to automatically turn customer emailed requests for quote (RFQs) to a metal service center into structured quotes to send back to the customer. This will enable salespeople to focus on selling and talking with customers, giving them the ability to respond with the single click of a button to “Send Quote”, rather than diving into the ERP to get the necessary information requested to respond to an RFQ.

Simple Code to Implement

1. Given the contents of an email, use your favorite AI tool(s) to determine if the email is an RFQ or not. Assuming the email is an RFQ, again use AI to extract the customer, products and quantities being requested, along with any other relevant information like custom dimensions or processing, due dates, shipping restrictions, etc.
2. Given this structured information from the RFQ, search the existing inventory to determine if the order can be fulfilled from products that are already in stock. If the order can either be fully or partially filled with existing stock, determine a reasonable sales price for the product.
3. Create a structured quote in the system for the customer with the calculated price, and build a corresponding UI for the salespeople to view a dashboard of all quotes, including these newly created draft quotes. Allow the user to finalize and “send” the completed quote to the customer.

Ideal Solution

The ideal solution will contain some form of the following components:

• Some UI to view the inventory and corresponding product details
• Some UI to input the email details
• Some UI to view all of the AI-created “draft” quotes that have been created in the system
  • These should contain information about the email that generated the “draft” quote
• Some UI for the salesperson to review the AI-created “draft” quote and “send” it to the customer
• Some UI to view all of the “sent” quotes in the system

Deliverables

• Please invite [email protected] to your Github repository
• A README with all instructions necessary to get the project running locally
• An eng “spec” outlining your general approach to the problem, including design choices, architecture decisions, and any assumptions or tradeoffs made
• A Loom or other recorded demo of your solution and its functionality

Evaluation Criteria

Your submission will be evaluated based on the following criteria: