GTM AI Operations Specialist

We are looking for a GTM AI Operations Specialist to design, build, and deliver AI agents that sellers use every day. This is a hands-on role focused on writing prompts, configuring agents, testing outputs, and launching seller-facing tools that reduce friction and help close revenue faster. 

This role connects Sales Operations, AI tooling, and enterprise data to put working AI agents directly in front of sellers. You will turn GTM priorities into seller-facing agents and automated workflows using platforms like Salesforce Agentforce, Microsoft Copilot Studio, Claude Code, Gemini, Snowflake, and Databricks. 

Key Responsibilities 

Prompt Engineering & AI Agent Development 

• Design and deploy autonomous agents: Prototype, test, and ship AI agents and custom copilots that handle key seller motions, including deal inspection, account research, opportunity updates, and meeting prep. 

• Optimize prompts: Write, refine, and maintain prompt templates, system instructions, and context that give sales teams accurate, consistent, and relevant output. 

• Tune agent behavior: Decide how agents interact with sellers and prospects. Set the guardrails, tone, persona, and fallback logic across platforms like Salesforce Agentforce and Microsoft Copilot Studio. 

Workflow Automation & Execution 

• Ship production workflows: Re-engineer revenue processes by connecting LLMs and agentic workflows to our core GTM systems such as Salesforce, Clari, Zoominfo, People.ai, Gong. 

• Prototype and iterate quickly: Build fast, test in the field, and refine using tools like Microsoft CoPilot, Claude Code, ChatGPT Enterprise, Gemini, and orchestrators such as Power Automate, or direct API integrations. 

• Build for scale: Document agent logic, prompt structures, and standard procedures so the automations you build can grow across the global GTM organization without depending on one person. 

Experimentation & Performance 

• Test prompts and models: Run structured experiments on system prompts, context retrieval, and model selection to improve output quality, seller adoption, and deal velocity. 

• Monitor live performance: Track, audit, and improve agent behavior and prompt accuracy in production, and fix output drift or performance gaps as they show up. 

Data, Enablement & Governance 

• Use enterprise data: Work with the Data and BI teams to tap into Snowflake and Databricks so agents can draw on real-time customer and pipeline data. 

• Drive field adoption: Build practical prompt libraries, quick-reference playbooks, and short training sessions that show sellers how to get real value from these tools. 

• Keep it responsible: Make sure every prompt, workflow, and data pipeline meets our security, privacy, and responsible AI standards. 

Required Qualifications 

• 3 to 5 years in Sales Operations, RevOps, GTM strategy, or AI operations, with hands-on experience putting prompt-engineered workflows or LLM agents into production. 

• Proven ability to write complex prompts and build agentic workflows on enterprise AI platforms such as Salesforce Agentforce, Microsoft Copilot Studio, Claude Code or Claude for Work, Gemini Enterprise, or ChatGPT Enterprise. 

• Strong working knowledge of Salesforce and comfort with enterprise data environments like Snowflake and Databricks. 

• A builder's mindset. You enjoy testing, troubleshooting, and shipping working automations rather than just putting together slide decks. 

• Comfortable using data through SQL, Power BI, or spreadsheets to track adoption, agent performance, and productivity gains in the field. 

Preferred Qualifications 

• Experience with revenue intelligence tools like Clari, Gong, People.ai, and other GTM tools. 

• Relevant certifications are a plus, including Salesforce Agentforce, Microsoft Copilot accreditation, Azure AI Fundamentals, or recognized prompt engineering or enterprise AI certifications. 

Core Competencies 

• Prompt engineering and agent design 

• GTM execution and workflow automation 

• Working knowledge of Salesforce, Snowflake, and Databricks 

• Rapid prototyping and experimentation 

What Success Looks Like 

• Sellers spend measurably fewer hours a week on admin work. 

• Agentforce, Copilot, and custom prompts see strong, active use across the field. 

• Live prompts and agents run with low error and drift rates. 

• A standardized prompt library and GTM agent catalog are up and actively used.