Turn Conversations Into Actionable CRM Data With the In-Person Meeting Assistant

Capturing insights from in-person meetings can be a challenge. Teams often rely on ad hoc transcription apps or personal note-taking to document conversations. That means important context stays outside of Salesforce — and admins are left trying to bring consistency and governance to information that never made it into the CRM in the first place.

For Salesforce Admins responsible for system architecture and data governance, this creates a familiar challenge: How do you capture sensitive, face-to-face conversations in a way that’s secure, compliant, and consistent across your organization? Without a standardized approach, teams may turn to disconnected tools that create data silos and introduce unnecessary risk. I’m happy to announce that In-Meeting Transcription is now available in the Salesforce mobile app as part of the 260.030 release, solving this exact challenge.

The In-Person Meeting Assistant is a new on-device feature that captures and transcribes face-to-face conversations directly within the Salesforce mobile app. It helps teams document meetings in a secure, standardized way that keeps conversation data connected to your CRM.

This mobile-native intelligence is built with user experience and admin governance in mind.

• Frictionless start: Users are prompted with a pre-meeting notification before the event to quickly launch transcription, ensuring no insight is missed.
• Uninterrupted workflow: With support for iOS Live Activities (including Dynamic Island) and Android Live Updates, transcription continues seamlessly in the background, even when the phone is locked or users switch to other apps.
• Intelligent tagging: Post-meeting, the app automatically identifies different speakers, helping reps quickly assign names with intelligent accuracy before syncing to Conversation Intelligence (previously known as Einstein Conversation Insights).
• Mandatory consent gate: Your compliance is protected by a required pre-transcription consent screen that ensures users secure verbal agreement from all participants.
• Privacy-first architecture: This security model is built on on-device processing, meaning audio data is never stored in the cloud or leaves the device, providing superior data residency and mitigating risk.

With the right governance in place, admins can transform in-person conversations from scattered notes into trusted CRM data that powers automation, reporting, and AI-driven insights. Once the meeting is transcribed, get access to call summaries or dive into detailed transcripts, and automatically surface mentions of pricing, products, customer objections, and more directly in call records.

Your governance and rollout checklist

The In-Person Meeting Assistant is a privacy-first, on-device transcription engine, but its success starts with your governance decisions. You define the compliance standards for face-to-face intelligence. Before you turn it on, the key decisions for rollout involve:

• Security and data residency: All transcription and speaker identification happens locally on the device. No audio data is ever stored or sent to the cloud for processing, and upon user logout, all local transcripts from the Salesforce mobile app are immediately wiped. This mitigates the highest risk associated with third-party, cloud-based audio solutions.
• Compliance and consent: As the architect of conversation data, you control the entire data lifecycle. This means not only defining the mandatory consent workflow — where reps check a box confirming they’ve obtained verbal consent from all participants — but also establishing data governance policies, including retention settings for transcripts once they’re synced to Conversation Intelligence. Your change management plan must ensure reps understand their compliance obligation to secure audible consent before they enable the digital consent gate.
• Permission management: You, the admin, are the gatekeeper defining which users can access and leverage this new data stream. The feature is governed by specific user permissions for Conversation Intelligence. You must audit permission sets to ensure only authorized reps can start a transcription session, thereby governing who can inject sensitive face-to-face data into your CRM and ensuring proper data stewardship across your mobile workforce.

As one CIO at a major state agency recently shared, “Our field investigators are out interviewing people involved in sensitive cases. They need to capture those conversations accurately and get them into the CRM immediately without risking cloud-based audio storage.”

This technical constraint (On-Device ONLY) is your primary strategic advantage in consolidating risk and eliminating shadow transcription tools across your organization.

Turning conversation data into automation

Capturing the transcript is only the beginning. The real magic happens once that transcript syncs to Conversation Intelligence. This is where you design the automation that converts raw conversational data into actionable business intelligence.

Once the data is in Conversation Intelligence, you can:

• Automate summaries: Use Einstein to generate concise meeting summaries and action items automatically.
• Access automatic insights: See out-of-the-box insights related to Next Steps, trending terms, and customer objections directly on call records. You can also configure keyword insights about the competitors, products, and custom keywords relevant to your teams.
• Build smarter flows: Use invocable actions to trigger a Slack notification or a follow-up Task based on specific keywords found in the transcript.
• Fuel Prompt Builder: Ground your generative AI prompts in real conversation data to create hyper-personalized follow-up emails.
• Visualize in Data 360: Pull meeting insights into Data 360 to see how face-to-face interactions impact your overall deal velocity.

Why this matters for admins

By implementing this solution, you position yourself as the strategic leader unifying conversational intelligence. You’re not simply deploying a feature; you’re defining the new enterprise architecture for face-to-face data.

Why this matters strategically

• Standardization: This tool ensures conversation data is captured uniformly across the organization, feeding reporting, forecasting, and AI responsibly.
• Mitigate risk and consolidate tech debt: You eliminate costly, disjointed third-party transcription apps and consolidate risk by removing shadow transcription tools.
• Architecture fit: The data syncs directly to Conversation Intelligence, making conversation data an organic part of your Salesforce architecture.

Ready to get started?

If your teams rely on in-person meetings, this feature gives admins a secure way to capture those conversations and bring them directly into Salesforce. 

The In-Person Meeting Assistant is scheduled for general availability starting early April 2026.

Start by enabling In-Person Meetings in Conversation Intelligence, reviewing the required permissions, and defining your change management plan for consent in the transcription flow, to ensure you can convert raw transcripts into operational momentum. Your data governance — and your end users — will be better for it.

Resources

• Salesforce Help: Set up Mobile AI Transcription
• Trailhead: Einstein Conversation Insights for Sales
• Vidyard: Coach Reps and Drive Revenue with Einstein Conversation Insights
• Salesforce Help: Einstein Conversation Insights