Adding AI Capabilities to Salesforce

Initially, we imported core sales plays into Salesforce and standardized their templates. A foundational library was created for Customer Challenges (CCs) and Sales Plays (SPs). Einstein AI was then trained on datasets including user notes, phone transcripts, CRM activity logs, campaign engagements, and customer intent signals like web activity and event attendance.

The AI provided immediate, "on-rails" recommendations without requiring manual user intervention. Users could review a full, ranked library of suggested challenges and plays directly within the Opportunity page, significantly streamlining their workflow.

We evolved the feature by introducing a conversational AI experience through Einstein Copilot, enhancing user interaction and refining the front-end experience on the Opportunity page. Einstein Copilot proactively engages sales reps with targeted questions, clarifying customer context around pain points, business goals, and strategic focus areas such as growth, compliance, or efficiency.

To ensure high accuracy and relevance, Einstein AI leveraged a diverse set of variables including:

• Deep web searches on companies for news and historical performance
• Industry-matched opportunities and previous renewal data
• User interaction data (Changes to account details, notes, support cases...)
• Direct conversation logs from Einstein Copilot

Throughout design and development, I collaborated closely with cross-functional teams, including stakeholders, solution architects, and developers. Continuous feedback loops through iterative testing informed feature improvements, ensuring usability and efficiency.

As a result, Salesforce teams reported significantly increased productivity, more consistent messaging, and greater confidence in sales play selection.