Hi everyone,
I'm exploring a solution where sales reps receive customer requirements as images, PDFs, Excel files, etc., and manually create Quotes/RFQs in Salesforce. Validation currently depends on 20–25 reference documents (pricing sheets, catalogs, spec sheets) stored on engineers' local machines, which sometimes leads to incorrect quotes.
I'm looking to use Agentforce/AI to:
- Identify products/parts from customer documents.
- Validate Quotes/RFQs against the customer's original requirements and flag mismatches.
- Suggest quote line items automatically.
Constraints:
- Prefer a configuration-first approach with minimal custom development.
- Quote-related documents are stored as Salesforce attachments/files, so the agent should be able to analyze them during validation on individual opportunities and related RFQs.
Has anyone implemented something similar? I'd love to hear your experience with Agentforce, Data Cloud vector search/RAG, document ingestion, and where you'd draw the line between Salesforce-native capabilities and custom AI solutions.
Also, if you have any recommended articles, blogs, or learning resources on this topic, I'd really appreciate them.
Thanks!
It sounds like you first need to streamline the process so that customers use a standardized way of sending their requests, then think of any kind of CPQ implementation to sort out the product catalog, and only when it's all done and clean, add AI.
