NOTE: Demo visuals use either blurred real data or synthetic placeholders to protect customer privacy.

While early versions streamlined manual research, new challenges surfaced in real-world usage:

• Partner-scheduled meetings introduced ambiguous event details.

• Reps could not always identify the correct prospect account or shared opportunity.

• Conversation starters lacked accuracy without reliable location or persona data.

This update introduced a workflow that accurately identifies prospect versus partner accounts based solely on emails and calendar entries, linking them to the corresponding shared opportunity. The result: targeted research delivered even when meeting inputs are messy.

• Detects the accurate prospect account from partner-involved invites.

• Surfaces opportunity-level analysis tied to the correct account.

• Expands activity view with personas, personal context, and engagement data.

• Enhances “create a connection” with Salesforce location intelligence.

• Improves recent news feed for both general and cybersecurity updates.
  

As Haren explained: “From the emails and from the event title, [the bot] can pick up exactly which shared opportunity is being looked at to provide pinpoint research.”

The team also refreshed UI elements and iterated quickly based on pilot feedback. One pilot user noted that the improved conversation starters were especially helpful.

This iteration focused on precision, ensuring reps can trust the prep output in complex scheduling scenarios. Early signals from pilot users indicated greater confidence.

• More accurate mapping of meetings to the right opportunities.

• Higher relevance in conversation prompts using Salesforce data.

• Sustained time savings of ~3 hours per rep per week.

• Pilot users cite faster comprehension of customer context.

Next step: expand pilots beyond the initial 12 users and continue integrating feedback loops into the bot’s workflows.

Pilot feedback signaled that the tool crossed a trust threshold. Users no longer viewed it as an experiment, but rather as a reliable part of their workflow. That cultural shift (from trying AI to depending on it) is as impactful as the hours saved.