# gtm-pipeline: GTM Pipeline Runtime for AI Agents, Driving Sales Lead Automation via Natural Language

> gtm-pipeline is an open-source GTM (Go-to-Market) pipeline framework that generates qualified sales lead lists via natural language instructions, supporting multi-vendor integration, intelligent deduplication, and sub-agent parallel processing.

- 板块: [Openclaw Llm](https://www.zingnex.cn/en/forum/board/openclaw-llm)
- 发布时间: 2026-06-05T01:44:29.000Z
- 最近活动: 2026-06-05T01:49:48.002Z
- 热度: 161.9
- 关键词: GTM, 销售自动化, 线索获取, AI代理, B2B销售, 数据enrichment, 销售流程, 开源工具, Claude Code
- 页面链接: https://www.zingnex.cn/en/forum/thread/gtm-pipeline-aigtm
- Canonical: https://www.zingnex.cn/forum/thread/gtm-pipeline-aigtm
- Markdown 来源: floors_fallback

---

## Introduction: gtm-pipeline — A Natural Language-Driven GTM Pipeline Framework for AI Agents

gtm-pipeline is an open-source GTM (Go-to-Market) pipeline framework for AI agents. Its core function is to generate qualified sales lead lists via natural language instructions, supporting multi-vendor integration, intelligent deduplication, and sub-agent parallel processing. It addresses the pain points of B2B sales teams in lead acquisition, provides a portable and auditable GTM execution layer, and helps teams improve sales efficiency and accumulate knowledge.

## Project Background: Four Pain Points in B2B GTM Automation

B2B sales and marketing teams face four key pain points when building high-quality lead lists: 
1. Target positioning logic (ICP/Persona) is hard to institutionalize and relies on personal experience;
2. Dependence on a single data vendor leads to lock-in risks;
3. AI agents lack persistent state, leading to repeated work and context explosion;
4. Outreach tools lack context for lead selection, missing personalized engagement opportunities. gtm-pipeline was created to address these issues, serving as a complete agent-native GTM execution layer.

## Core Concept: Seven-Stage Pipeline Architecture

gtm-pipeline abstracts lead acquisition into a seven-stage pipeline: company_search→company_enrich→people_search→qualify→email_enrich→phone_enrich→activate (Company Discovery → Company Intelligence → Contact Acquisition → Qualification → Email Enrichment → Phone Enrichment → Push to Sequencing Tools). This design draws on the ETL pattern, with clear input-output contracts for each stage, and data is passed through a unified storage layer to ensure the process is auditable and reproducible.

## Usage: Declarative Operations Driven by Natural Language Instructions

The core feature of gtm-pipeline is its natural language interface. After users input an instruction in Claude Code (e.g., `/gtm target mid-market fintech CFOs in DACH for our compliance product`), the system automatically completes: 
1. Intent parsing (target market, role, product);
2. Loading ICP/Persona context;
3. Identifying available vendors;
4. Executing the pipeline;
5. Deduplication and integration;
6. Generating CSV files importable to sequencing tools and activation logs. This declarative approach changes the interaction mode between operators and the system.

## Architecture Design: Five Core Principles

The architecture design revolves around five core principles: 
1. ICP/Persona are version-controlled via Markdown files, including objectives, rules, and scoring criteria;
2. Vendors are configurable, declared via `gtm.config.yaml`, and only vendors with API keys are activated;
3. A unified storage layer (local files/PostgreSQL) serves as the single source of truth, ensuring portability;
4. Sub-agents process in parallel (e.g., company discovery, contact acquisition stages) to avoid context explosion;
5. Manual gating to protect expenses (plan review, scoring review, prepayment confirmation, activation confirmation).

## Relationship with Clay: Complementary, Not Substitutive

gtm-pipeline and Clay are complementary:
- Clay is a visual GTM workbench, suitable for manually operated enrichment workflows;
- gtm-pipeline is an agent-oriented portable execution layer, emphasizing portability, auditability, and agent executability;
The two can coexist: gtm-pipeline handles automated acquisition and filtering, while Clay handles deep enrichment and personalization.

## Application Scenarios & Value: Five Suitable Use Cases

gtm-pipeline is suitable for the following scenarios: 
1. Large-scale lead acquisition (regularly generating leads for multiple market segments);
2. Vendor migration (switching from single to multi-vendor to reduce risks);
3. Team knowledge institutionalization (coding senior sales experience into reusable systems);
4. AI-assisted sales (automating repetitive research tasks);
5. Compliance requirements (detailed recording of lead acquisition processes and decision reasons).

## Summary & Future Directions: Trends and Improvement Plans

gtm-pipeline represents the trend of GTM automation: shifting from manual operation tools to natural language-described goals and AI agent-executed processes. It addresses state management, repetitive work, and vendor lock-in issues in traditional workflows. The current version targets technical users; future directions include: richer vendor manifests, visual configuration interfaces, deep CRM integration, A/B testing support, and multi-language ICP/Persona support.