The Challenge
Ramp moves $100B+ in annual spend for 50,000+ companies. You're not automating marketing with AI-you're building autonomous agent fleets that run 24/7, learn from outcomes, and own the entire marketing operation end-to-end.
Your Mission
Ship first autonomous agent handling one high-volume marketing workflow (content generation, lead enrichment, or campaign ops) with measurable quality evals
Define and document agent decomposition framework: skills, tools, evaluation criteria, memory patterns, and orchestration layer for marketing use cases
Build evaluation system that catches regressions and validates agent output quality without manual review bottlenecks
Establish baseline metrics on agent autonomy score, error rates, and cost per output vs. manual workflow
Deploy 3-4 autonomous agents operating in production across content, campaign, and channel optimization workflows with <5% manual intervention rate
Implement self-improving loops where agents monitor outputs, flag quality issues, and improve without human intervention cycles
Create internal agent infrastructure and documentation enabling marketing team to operate, trust, and iterate on agents independently
Measure impact: track agents' contribution to campaign velocity, cost reduction, and revenue lift vs. pre-agentic baseline
KPIs You'll Own
Agent Autonomy Score
Percentage of workflow tasks completed without human intervention or approval loops per agent.
Output Quality / Regression Rate
Error rate and quality degradation detected by evaluation framework; target <2% regression week-over-week.
Cost per Output
Fully loaded cost (compute, API calls, human oversight) per completed marketing deliverable vs. baseline manual cost.
Workflow Cycle Time
Hours from trigger to completion for agent-operated workflows; measure acceleration vs. manual process.
Agent Self-Improvement Rate
Measurable quality or efficiency gains driven by agent feedback loops and learning, excluding manual retraining.
Tools & Stack
Your Team
Your Manager
Head of Growth Marketing or VP Marketing (not specified; clarify in initial conversation)
Current Team
Ramp marketing team already deeply leveraging AI; you're scaling from AI-assisted to agent-operated
New role; building out agentic AI capability
The Package
Salary
$160K-$200K base
Remote
On-site, New York, NY
Benefits & Perks
Company Intelligence
Ramp is fintech infrastructure for enterprise finance: automating how companies authorize payments, flag risk, categorize spend, and close books. 50,000+ customers manage $100B+ in annualized spend on the platform. Ramp customers save 5% and grow revenue 16% in year one.
Team Size
500+
Customers
50,000+ companies
Culture
High agency, high urgency. Slope over intercept. Care less about where you trained, more about what you've built. Everyone owns problems end-to-end and makes consequential decisions.
Is This Role For You?
- You've shipped AI/ML projects from prototype to production and care deeply about eval frameworks and feedback loops
- You're extremely AI-pilled: you've read the papers, run benchmarks, built multi-agent systems, and have strong opinions on tool-use vs. code-gen and RAG vs. fine-tuning
- You think in systems and loops, not tasks. You own end-to-end: from identifying which workflows to automate, to building, deploying, monitoring, and iterating
- You can scope an MVP, mock it, and ship it yourself. You're equal parts PM and engineer with the conviction to make autonomous systems work
- You're looking for a pure AI research role or want to stay on the frontier without shipping to production constraints
- You need flexibility: this is on-site NYC, full-time, and agent operations require being embedded with the marketing team
- You expect someone else to define the problem or hand you a roadmap; you have to own discovery, scoping, and outcomes
Interview Process
Initial conversation
Discuss your agentic AI experience, past agents you've built, and your opinions on tool-use, evals, and orchestration patterns.
Technical deep-dive
Walk through a past project: how you decomposed workflows, built evals, and measured agent quality. Focus on systems thinking and feedback loops.
Design exercise
Given a marketing workflow (e.g., lead enrichment, content production), sketch out the agent architecture you'd build: skills, tools, evals, memory, and orchestration.
Team conversation
Chat with marketing team about how you'd operationalize agents and enable non-engineers to trust and iterate on them.
Ready when you are
Interested in this role?
Apply now and hear back within days, not weeks.
Get alerts for growth marketing jobs
Weekly email. Unsubscribe in one click.
Keep exploring
Related searches
More like this
Roles similares
Product Marketing Lead, Market Growth & Expansion
Waymo • Mountain View, CA
Senior Manager, Digital Marketing
PWX Solutions • New York, NY 10018
Growth Marketing Manager
New U Women’s Clinic & Aesthetics • Kennewick, WA 99336
Senior Marketing Strategy & Planning Manager
Rippling • New York, NY 10007
Context
About Growth Marketing roles
Growth marketers drive user acquisition, activation, and retention through data-driven experimentation. They sit at the intersection of product, data, and marketing — running A/B tests, building funnels, and scaling what works.