Meeting Preparation with AI Chat

Workflow recipe: Meeting Preparation with AI Chat

AI Chat is a conversational interface to your Affinity CRM: ask questions in natural language and get answers grounded in your own data, with cited sources. At GA it can also take confirm-before-execute actions — so you can move from “what do I know about this meeting?” to “update the record” without leaving the conversation.

Asking vs. the automatic brief

Two complementary ways to get meeting context:

• The Meeting Prep agent produces your pre-meeting brief for the day’s external meetings — generated overnight and ready by 7:00 AM in your local time zone, so you don’t have to ask. It’s delivered in-app: on your home page (Start Your Day / Ready for You), in the Ascend Agent Activity feed. This is the reliable way to get a full pre-meeting brief. (See Meeting Prep & Data Update.)
• Asking AI Chat (this article) is on-demand and conversational — explore what your CRM knows about a specific company or contact right now, dig deeper with follow-ups, and act on what you find. Reading your Meeting Prep brief in Chat, and reviewing Suggested Field Updates in Chat, are fast-follows after launch — not part of GA. The Meeting Prep and Data Update agents run automatically on their own; the meeting-prep prompt below works as a normal Chat query.

Before you start

• Plan: AI Chat with actions is available on Scale, Advanced, and Enterprise.
• Access: AI Chat is GA. Admins control agents and AI features in Settings → Automations & AI → Affinity AI; if you don’t see AI Chat, check with your admin.
• Web app only. Affinity AI only sees and can change what your own account already can.

Step-by-step

1. Start from the meeting (fastest way)

• Open Meetings in the left nav.
• Use the ← / → arrows to move to the meeting’s date.
• Open the external meeting entry.
• From there you can pick the company profile and add a meeting note.
• Click Prep for meeting — this launches AI Chat with a meeting-prep prompt prepopulated from that company and meeting, so you don’t have to type anything.

You can still type a prompt manually instead — Prep for meeting just does the first step for you.

2. Open AI Chat and ask for a prep

Help me prep for my meeting with [Company]

Affinity AI shows a brief thinking state, then a short “sources used” line.

3. Read the response

Typically: relationship history, recent activity (last interaction and how recent), and open items (deal/pipeline context). Review it against the cited sources.

4. Ask follow-ups

When did we last talk to them, and who was on the thread?
What's the current pipeline stage for this deal?
Summarize the last note on this company.

5. Act on it — with your approval

When ready, ask Affinity AI to take an action right in chat. At launch: update a field or pipeline stage, add to a list, create a note, create a reminder — for example:

Move [Company] to [stage]
Log a note on [Company] summarizing this prep
Remind me to follow up with [Company] next Tuesday

Affinity AI shows what will change and waits for you to confirm. Nothing is written until you approve — the same confirm-before-execute model used across all Ascend agents. A few actions (upload a file, saved view, create a list) come later.

Tips

• Name the company or person explicitly. Person search is by name or email today (attribute search is post-launch).
• Verify dates against your calendar — occasionally a past meeting may show as “upcoming.”
• Use follow-ups instead of one giant question.
• Review before you approve — check the proposed change against the cited source.

Common questions

Will Affinity AI change anything on its own? No — for any write action it shows the proposed change and waits for your approval.

What can it do at launch? Answer questions (with cited sources) and take four confirm-before-execute actions: update a field or pipeline stage, add to a list, create a note, create a reminder.

What if the answer isn’t quite right? Ask a follow-up or check the cited sources. Note the limits: ~90-day activity window (being extended); person search by name/email today; no location filtering yet; no reliable point-in-time field history.

---

Affinity AI can make mistakes. Review each suggestion before approving it.