ye myat
book intro →

AI Agents for Pre-Call Briefs

Walk into every client call already knowing what happened last time, what's likely to come up, and what to ask. Prepared automatically the night before.

01the problem

You have a client call in 10 minutes. You scramble to find last week's notes, scan Slack for anything you missed, and try to remember what you promised. Half the time you walk in underprepared. The other half, you spend 15 minutes before each call doing research that could have been automated.

02what the agent does

Every night at 10 PM, the agent checks tomorrow's calendar for external meetings. For each one, it pulls recent emails, Slack threads, meeting notes, and project tasks involving that person. Then it writes a brief: what the meeting is about, three things likely to come up, open items to resolve, and one smart question to ask. You wake up prepared.

Example output

agent output

Pre-Call Brief — Acme Corp Q1 Review — 10:00 AM

Attendees: Sarah Chen, Jake Torres

What this meeting is about Quarterly review of the product launch campaign. We shipped the landing page last week and ad creative is in final review. This is likely a checkpoint before the March 25 launch date.

Three things likely to come up

  1. Ad budget approval (Sarah asked about this in Slack on Mar 4)
  2. Landing page conversion tracking setup (open task, due this week)
  3. Post-launch measurement plan (mentioned in last call but no follow-up yet)

Open items to resolve on this call

  • Final sign-off on ad creative variants (sent for review Mar 5)

One question to ask

What does a successful launch look like for your team beyond the metrics we're tracking?

Recent context

  • Last interaction: Mar 5 — Sent ad creative for review via email
  • Relevant news: Acme announced Series B funding on Mar 3

Try it yourself

Copy these prompts into Claude Cowork. Connect your tools. The agent handles the rest.

01
Scheduled pre-call briefs
Runs every night. Briefs for all tomorrow's external meetings land in your inbox.
Run every weekday at 10:00 PM.

1. Pull tomorrow's calendar events
2. Filter for meetings with external attendees
3. For each meeting, search last 14 days of: emails with attendees, Slack mentions, meeting notes, related project tasks
4. Write a brief: what it's about, 3 things likely to come up, open items, one question to ask, recent context
5. Save to projects/[client]/call-prep/YYYY-MM-DD-[name].md
6. Email all briefs to me. If no external meetings tomorrow, email: 'No external meetings tomorrow.'

FAQ

Common questions about setting up this agent.

How does the agent know what's likely to come up?+
It reads recent context: emails exchanged, Slack conversations, previous meeting notes, and open tasks. If Sarah asked about ad budgets in Slack three days ago and there's no resolution, the agent surfaces that as something likely to come up. It's pattern matching on your actual communication history, not guessing.
Can I customize the brief format?+
Yes. The format shown here is a good default, but you can adjust it. Some people want a shorter version (just open items and one question). Others want competitor intel included. You modify the prompt once and it applies to all future briefs.

More for agencies

Scope Creep Detection

Automatically compare what was agreed against what's actually being delivered. Catch scope creep before it eats your margin.

view

Client Reporting

Generate weekly client reports automatically from your project tools, Slack, and meeting notes. No more Friday afternoon scrambles.

view

Retainer Health Monitoring

Track retainer usage, spot disengaged clients, and catch churn signals before it's too late. Runs monthly on autopilot.

view

Want this running for your team?

I set up AI agents like this for agencies. 30 minutes, no strings.

book intro →