Skip to content
GTM Agent Weekly

Fill most CRM deal fields from calls and answer deal questions in Slack

After every sales call, this agent fills in the key deal fields in the CRM, like pain points and next steps, and answers deal questions in a chat channel. Reps stop retyping notes and handoffs to customer success get easier.

Paste this into ChatGPT, Claude or your agent tool. See the prompt

  • Pipeline Hygiene / CRM
  • intermediate
  • RevOps
  • AE
  • CS
Category
Pipeline Hygiene / CRM
Difficulty
Intermediate
Est. cost
Not stated by source

Works with

Problem

Reps used different note-taking tools, so call insights arrived in inconsistent formats and were hard to link to the right deal. Handoffs from sales to CS were manual.

Setup

A founding RevOps lead consolidated call recording into Attention, which extracts pain points, next steps, use cases and competitive intel from calls, links meetings to deals automatically, and writes them into CRM fields. A Slack channel with Attention's Super Agent lets anyone in the company, including people who never record calls, ask about deals and customers.

Prompt

Starter prompt by The GTM Agent Weekly, written from the setup described on this page. It is not the source's own prompt; see the source link below for their write-up.

Starter prompt · 27 lines
# ROLE
You fill CRM deal fields from sales call transcripts and answer questions about deals and
customers for anyone in the company, so reps rarely type into the CRM.

# INPUTS
- {{TRANSCRIPT}}: one call transcript with date, attendees and the recording link.
- {{CRM}}: deals, contacts and accounts. {{FIELD_LIST}}: the deal fields to fill, with a
  one-line definition and allowed values for each (e.g. pain points, next steps, use case,
  competitors mentioned).

# TASK A: AFTER EACH CALL
1. Match the call to its deal: attendee emails to contacts, then contacts to the open deal.
   If there is no single match, stop and post "Needs deal link" to {{REVOPS_CHANNEL}}.
2. For each field in {{FIELD_LIST}}, extract a value only if it was said on the call. Include
   a short quote and timestamp as evidence.
3. Write values to empty fields. If a field already has a value, propose the change instead
   of overwriting.

# TASK B: ANSWER QUESTIONS IN {{CHAT_CHANNEL}}
Answer from CRM records and transcripts only. Cite the call (date, link) behind each claim.
If you do not know, say so.

# GUARDRAILS
- Never fill a field from guesswork. Blank is better than wrong.
- Respect record permissions: do not reveal deals the asker cannot see. Treat deals marked
  {{SENSITIVE_FLAG}} as off-limits in shared channels.
- Weekly, report the share of required fields filled by the agent vs. by hand at {{STAGE}}.

Replace every {{PLACEHOLDER}} before running. Works with any capable LLM and runner.

Tools

Steps

  1. Consolidate call recording onto one platform so transcripts share a format.
  2. Define which deal fields the AI should extract and fill from calls.
  3. Turn on automatic meeting-to-deal association so calls land on the right record.
  4. Open a Slack channel for the agent so product, marketing and CS can self-serve answers.
  5. Check field completeness at the contract stage to see how much reps still enter by hand.

Results

Vendor case study reports that most required fields are already filled by contract stage; self-reported.

What goes wrong

Gotchas

  • Deals created after the meeting are hard to link to earlier calls unless association is automatic.
  • Company-wide Slack access to call insights raises data-access questions for sensitive deals (editor note).
  • Vendor case study, so outcome figures are self-reported (editor note).

Source and credit

www.attention.com ›

Jamie Chen, Lovable (Attention customer story) · vendor

Source dated 2026-10-11 · last reviewed 2026-10-11. Summarized in our own words; see the source for details.

Related.

Flag booked meetings where the prospect barely committed

When a meeting gets booked, this agent reviews the call and flags ones where the prospect barely committed. Managers can firm up or requalify those meetings before a salesperson spends time on them.

Works with any of NooksOrumOutreach +11

  • Pipeline Hygiene / CRM
  • intermediate
  • Reference setup

Catch duplicate leads and missing fields with an n8n Salesforce agent

This agent checks new and changed leads for duplicates, including the same company spelled differently, and cleans them up with narrow, careful steps. The CRM stays tidy without someone hunting duplicates by hand.

Works with SalesforceSlack

  • Pipeline Hygiene / CRM
  • intermediate

Clean the CRM weekly with an agent that proposes fixes you approve

Once a week, this agent reviews the CRM for duplicates, bad values, gaps and dead deals and proposes fixes that a person approves before anything changes. Cleanup becomes routine instead of a rushed purge.

Works with HubSpotSalesforceSlack +1

  • Pipeline Hygiene / CRM
  • intermediate