Draft personalized outbound, recaps and forecasts for sales
SDRs are booking 3x qualified meetings and AEs are shaving ramp by 50%+. Task: Cursor's Head of Enterprise Growth built an internal AI tool, ChatGTM, for a 400+ person sales org. It aggregates data from Salesforce, Gong, LinkedIn, and the web to generate morning digests, personalized outbound drafts, call prep briefs, and live Q&A support. Sales reps review, tweak, and send these AI-generated assets via Gmail or Outreach, maintaining human control over final communications.; tools: Cursor. Human role: The original post does not say how a person stayed involved.; limitations: The tool is proprietary and not commercially available. Implementation requires deep integration with CRM, communication, and data warehouse systems. The source does not detail the technical architecture, maintenance costs, or specific data privacy safeguards. Success depends on high-quality input data and rep adoption..
Tools
Published on Agent Work Atlas:
Overview
SDRs are booking 3x qualified meetings and AEs are shaving ramp by 50%+.
Workflow
Cursor's Head of Enterprise Growth built an internal AI tool, ChatGTM, for a 400+ person sales org. It aggregates data from Salesforce, Gong, LinkedIn, and the web to generate morning digests, personalized outbound drafts, call prep briefs, and live Q&A support. Sales reps review, tweak, and send these AI-generated assets via Gmail or Outreach, maintaining human control over final communications.
Reported result
The source claims SDRs book 3x more qualified meetings and AE ramp time decreased by 50%+. These metrics are attributed to the author's report of Cursor's internal results. No independent verification or detailed methodology for these figures is provided in the text.
Human role
The original post does not say how a person stayed involved.
Limitations
The tool is proprietary and not commercially available. Implementation requires deep integration with CRM, communication, and data warehouse systems. The source does not detail the technical architecture, maintenance costs, or specific data privacy safeguards. Success depends on high-quality input data and rep adoption.
Try it
I want to build a prototype of an AI sales assistant similar to ChatGTM for my team. My business context is [replaceable: e.g., B2B SaaS selling to mid-market]. I have access to [replaceable: e.g., Salesforce, Gmail, and a data warehouse]. Please perform a dry run for one specific account: [replaceable: Account Name]. 1. Aggregate recent news, usage spikes, and exec changes for this account from [replaceable: data sources]. 2. Draft a personalized outbound email tied to a specific signal, including a suggested landing page angle. 3. Create a pre-call brief highlighting potential objections based on recent public statements by the contact. Output the results in a structured JSON format with fields: 'digest', 'email_draft', 'call_brief'. Do not send any emails or update the CRM. I will review the output for accuracy and tone before we proceed to automate this for a small batch of 5 accounts. If any data source is inaccessible, list the missing inputs and ask for clarification before proceeding.
Source reports are not independently verified. Publication date is the catalog publication date, not the original article date.