During the 2026-04-03 Salesforce progress call, the team identified a need to automate proposal creation for Quarra. Current proposal templates are built in Google Docs and suffer from persistent formatting issues — text boxes overflow across pages, content overlaps, and the output requires significant manual cleanup before it can be sent to clients. The team agreed to evaluate Salesforce's native AI platform, AgentForce, as the primary vehicle for building a proposal generation tool that lives inside Salesforce.
Quarra's current proposal workflow relies on manually maintained document templates that are difficult to control:
Lincoln Durham raised the idea of uploading templates to an AI to handle formatting automatically. Mark Hope noted that while AI-generated proposals are already viable (he uses the approach personally), the key challenge is where that AI lives and how it integrates with Salesforce data.
"You probably don't want everybody running off to Claude to create proposals... we can make a little app that basically said, I want to create a proposal, and you just go to the app and put in the answers. Or pull the answers from Salesforce."
— Mark Hope
Rather than routing users to an external AI tool (Claude, ChatGPT, etc.), the team agreed to explore building a custom proposal app inside AgentForce — Salesforce's native AI platform. This would allow:
Using external AI tools ad hoc was ruled out for practical reasons:
- No standardization across users
- No integration with Salesforce data
- No audit trail or version control within the CRM
Proposal generation is one of four Microsoft-related integration goals currently blocked pending server access from Quarra's IT team. The others are Outlook integration, Microsoft Teams integration, and automated folder creation on opportunity creation. Once server permissions are granted, file/folder creation triggered from Salesforce becomes possible — which is a prerequisite for saving generated proposal documents to Quarra's drive.
See: [1]