For this project, I built an AI-powered SDR assistant for a SaaS company using Salesforce Agentforce Builder. The goal was to streamline the sales development workflow by automating three core functions: retrieving and summarizing CRM account data, qualifying inbound leads and identifying customer needs, and scheduling demo meetings with prospects.

I designed the agent around a central "Main Agent" that routes conversations to six specialized subagents — Lookup_CRM_Accounts, Lead_Qualifications, Opportunity_Recommendations, Schedule_Meetings, Ambiguous_Question, and Off_Topic — each with its own focused instructions. The main agent was prompted to behave as a concise, professional internal SDR assistant that avoids unnecessary clarification and defaults to using available CRM context.

The biggest challenges I faced were refining how the agent routed requests between subagents, tuning CRM data retrieval, reducing unnecessary clarification loops, and improving each subagent's specialization so it stayed focused on its domain without overlapping with others.