LangChain says its GTM Agent generated 1,540 sales artifacts
LangChain says its internal GTM Agent generated 1,540 sales artifacts over the last 12 weeks, saving sales representatives thousands of hours across their workflows. Deals where a representative sent one of the documents to a customer moved to a later stage 47% of the time, compared with 16% for matched deals without one.
- GTM Agent generated 1,540 sales artifacts over 12 weeks.
- 47% of deals with a sent artifact moved to a later stage, versus 16% of matched deals without one.
- The artifacts include account plans, proof-of-value documents, proposal decks and onboarding plans.
- A review rule cut one one-pager’s runtime from 18 minutes to 7 minutes.
- Managed Deep Agents includes primitives used by the workflow, according to LangChain.
An agent for sales documents
LangChain’s GTM Agent creates documents that sales representatives previously assembled by hand for each deal. The artifacts include customized account plans, proof-of-value documents, proposal decks and onboarding plans.
The company says the agent generated more than 100 artifacts per week over the last 12 weeks. Representatives used the system to produce 1,540 artifacts, while documents sent to customers accounted for 337 hours of saved representative time. LangChain also reports that 46% of artifacts were followed by a customer call within 14 days.
The system was added to an existing GTM Agent that researches new leads, drafts personalized outreach for representatives to approve in Slack and surfaces account signals such as expansion opportunities and deal risks.
Context matters more than prompt tuning
LangChain says most of the quality comes from the context given to the agent rather than from prompt tuning. That context includes the company’s sales process, document templates, stage rules, brand guidance and requirements for moving a deal forward.
The agent reads Salesforce records, product usage, call transcripts, risks, decision criteria and next steps. It also uses LangChain’s brand assets and style guide to produce documents that follow the company’s visual rules. The brand kit, including logos, color and type tokens, fonts, and style guidance, is built into the sandbox image.
LangChain built the workflow from skills and instructions, a sandbox with Office tools, a Slack interface, scheduled runs, memory and tracing in LangSmith. The company says Managed Deep Agents includes the sandbox, skills and memory needed to build a similar system.
A review loop across the sales cycle
GTM Agent maps standard artifacts to five stages of the sales cycle. A representative can request a document in Slack, or the agent can suggest one during meeting preparation when it identifies a missing step in an active deal.
To create an artifact, the agent resolves the account in Salesforce, reads its Google Drive folder, asks for missing information such as success criteria or an economic buyer, and retrieves a leadership-owned template. It fills the template in a sandbox, renders the file, reviews the resulting pages and fixes layout problems before handing the document back to the representative.
LangChain says an early version could repeatedly review pages that were already acceptable. A rule to ship after one review finds no real defect reduced the runtime for one one-pager from 18 minutes to 7 minutes. The finished file is converted to native Google Slides, Docs or Sheets, while the representative receives a PDF preview and a Slack link with flags for any remaining human edits.
For users, the approach shows how LangChain is applying its agent tooling to a specific business process. The company’s results support its claim that structured business context, reusable templates and workflow integrations are central to making sales-document automation useful.
LangChain: How our GTM Agent builds consistent, high-quality sales artifacts
