Getting access and turning it on
- Available on Scale, Advanced, or Enterprise. At launch (July 22) Ascend is off until an admin turns it on; from August 5 it’s on by default for organizations that haven’t opted out.
- Turning the agents on or off is an admin, org-level control at Settings → Automations & AI → Affinity AI (the Ascend Agents card). In addition, each user can enable or disable individual agents — Meeting Prep and Data Update — on the Ascend → Agents tab.
- Affinity Notetaker — the primary source for post-meeting updates — is available on Scale, Advanced, or Enterprise.
The Meeting Prep agent — a pre-meeting brief
- The Meeting Prep agent scans your calendar and prepares a brief for your external meetings.
- Each brief pulls from your Affinity data: relationship history, the last interaction and how recent it was, who on your team is engaged, deal and pipeline context, and recent news on the company.
- Briefs are generated overnight and ready each morning by 7:00 AM in your local time zone. A brief surfaces in-app on the Affinity home page — in the “Start your day” / “Ready for you” modules — and in the Ascend Agent Activity feed. It appears in Affinity on the web (not email, Slack, or mobile). Open it with View Brief.
- Opening a brief brings up the Meeting Brief modal, which summarizes the person and relationship context and has two actions: Add as note — turn the brief into a note (it appears in Agent Activity as View Note; use this if you don’t have Affinity Notetaker, which otherwise creates that note automatically after the meeting) — and Archive, which keeps the brief in Agent Activity until you archive it, then moves it to Activity History.
- You get the same context you’d get by asking AI Chat “help me prep for my meeting with…,” prepared for you in advance so you don’t have to ask.
The Data Update agent — suggested field updates
The Data Update agent keeps your CRM current. At launch it works from your meeting notes.- Once a meeting note is created (Affinity Notetaker does this automatically; Granola or another integrated notetaker also works — see Does it need a notetaker? below), the agent reads the note and proposes CRM updates — usually in under a minute. At launch it triggers on meeting-note creation only.
- In the Agent Activity item, expand Review. Each row shows the field, its scope (“All Lists” or a specific list), the Current Value (with its last-updated date), and the Suggested Value. The source is the linked View Note.
- Select which fields to update via checkboxes (or Select all), then click Apply Updates. Leave a checkbox unchecked to skip that field. Up to 10 suggestions can appear per entity.
- It can suggest updates to any editable field — global or list-specific — including dropdowns and Ranked Dropdown fields (Status, Stage, Priority); for multi-value fields it appends rather than overwriting. (This is a step up from the Labs version of Suggested Field Updates, which only covered a subset of fields.)
- Nothing is written to the CRM until you click Apply Updates. Editing a suggested value before applying is not available at launch.
How the suggestions are made
When a meeting note is created, the Data Update agent picks it up automatically (new notes on external meetings). It works out which companies, people, and opportunities the note is about — from the note’s associations, or the meeting’s external attendees as a fallback — and which lists they’re on. A single meeting can produce suggestions for several records at once. It works from the records linked to the meeting — not other companies merely mentioned in the note (covering mentioned companies is planned for after launch). It then looks only at the fields you can already edit, reads the note against them, and proposes values it can support — labeling each stated (explicitly said) vs inferred. Dropdown suggestions must match your real existing options; sensitive fields like job title are only suggested when explicitly stated; and it skips anything ambiguous, unresolvable, or already correct. Nothing is written automatically.Where they live
- Find both agents under Ascend in the left nav and on the home page. Agent names appear on the Agents tab.
- Agent Activity — the queue of briefs and suggestions waiting for review. Activity History — what the agents have already done (applied vs skipped, with Previous → Suggested). Agents — turn each agent on or off (each is a background Monitoring agent).
- Note: Notetaker feeds these agents but isn’t part of Ascend.
How to stay in control
- Review before it’s real. Updates wait in Agent Activity until you act — see Review & approve an agent’s suggestions.
- It shows its work. Every suggestion names the field, its scope, and the Current → Suggested value, and links to the source via View Meeting Note.
- Quality-filtered. Low-quality, ambiguous, or already-correct suggestions are filtered out before they reach you, and no more than 10 appear per entity.
- It won’t overwrite newer data. If a field’s value has changed since the suggestion was generated, that suggestion is rejected on apply.
- It won’t fire your automations at launch. Applying a Suggested Field Update does not trigger Status/Stage automations at launch.
- It respects your permissions. Each agent only sees and changes what your account already can.
Common questions
- Does it need a notetaker? It works from your meeting note, so a note has to exist. Affinity Notetaker is the primary source — it takes the note and attaches it to the meeting automatically. Granola and other integrations also work, but historically they attach notes to the person/company record only, not the meeting — and the agent triggers off the meeting note (so those won’t trigger updates). See Getting your meeting notes into Affinity.
- Will it change my pipeline automatically? No. Updates queue for your approval; nothing applies on its own.
- Where do the pre-meeting briefs show up? In-app on the “Start your day” / “Ready for you” homepage modules and the Ascend Agent Activity feed. Affinity web only. Open one with View Brief.
- What if it suggests the wrong value? Leave its checkbox unchecked so it isn’t applied. Editing a suggested value before applying comes after launch.
- How is this different from the Labs “Suggested Field Updates”? The agent version covers every editable field, including list-specific fields and dropdowns; the Labs version only covered a subset.
- Which meetings does it cover? External meetings (internal meetings are excluded at launch).
Affinity AI can make mistakes. Review each suggestion before approving it.