About the Role

Most growth teams don't lack ideas. They lack the infrastructure to test them fast enough to matter.

We're hiring an Agentic Operator whose core output is one thing: reducing the time between hypothesis and signal. You'll build the agent pipelines, orchestration layers, and automation systems that let our growth team run more experiments in parallel — and learn from them faster — than any traditionally-staffed team could.

This is a builder role. You'll treat AI agents as a programmable workforce, replacing human bottlenecks in the experimentation loop with systems that pre-validate, instrument, execute, and surface insights automatically. You won't be handed a playbook. You'll write it in code. You'll report to the Director of Demand Generation and work closely with channel owners, lifecycle, product, and sales to make the growth team's experiments faster, more rigorous, and more evidence-based.

The right person is operationally sharp with genuine growth instincts. You understand B2B SaaS funnels well enough to prioritize experiments intelligently and catch flawed hypotheses before they waste the team's time. You're technically fluent enough to build AI workflows, instrument experiments, and own the tracking systems the team depends on. And you bring the organizational discipline to manage a high-velocity experiment backlog without letting things slip through the cracks.

What You'll Build

Agent Pipelines for Concept Pre-Validation

Before human resources touch a test, your agents have already scored it — LLM-evaluated copy variants, synthetic audience testing, funnel drop-off analysis. Only the strongest hypotheses make it to execution.

An Automated Experiment Queue

You own the full loop: intake, prioritization, instrumentation, and readout — all accelerated via automation. Funnel drop-off doesn't trigger a Monday meeting; it triggers the next test.

An Experimentation Process

You own the full loop — backlog, sequencing, instrumentation, and readout — with the rigor to match. Every test ships with a falsifiable hypothesis, clean measurement setup, and defined success metric. Null results get documented, not buried. Channel owners, lifecycle, and PLG stay in sync so the team is always running the highest-leverage tests without conflicts or duplication.

Stack Orchestration

You wire together our existing tools — Amplitude, HubSpot, Clay, REO.dev and enrichment layers — so signals flow automatically into action. You integrate frontier LLMs (Claude, GPT) and workflow tools (n8n, Gumloop) into production-grade systems that improve over time.

Parallel Experimentation Infrastructure

Instead of sequential tests, you build the infrastructure to run experiments across channels simultaneously — compressing the learning cycle and multiplying the team's throughput without multiplying headcount.

What You'll Bring

Growth Fundamentals

You understand acquisition funnels, CRO, CAC/LTV, and hypothesis-driven testing deeply enough to know which experiments are worth building for. Instinct for what moves the needle isn't optional — it's the filter on everything you automate.

Agentic Build Skills

You've shipped multi-agent workflows in production, not just personal automation scripts. You're fluent in API-based orchestration, prompt engineering, and building feedback loops that self-improve. You've built internal tooling that other teams depended on.

Systems Thinking

You design for leverage. Before touching a tool, you ask: what's the bottleneck, what happens if this scales 10x, and where does the loop close? You build things that get smarter over time.

Zero-to-One Comfort

The playbook for this role doesn't exist yet. You've operated in environments where you had to define the process, not just follow it — and you preferred it that way. You want the opportunity to build, improve, and adapt a new experimentation motion.

Multi-stakeholder Management

Experience managing or coordinating a multi-stakeholder experimentation program — not just running your own tests, but helping a team run better ones.

You Might Be a Fit If…

• You can describe a specific experiment cadence you drove — e.g., "we went from 3 tests/month to 18" — and walk through exactly what you automated to get there
• Your first instinct when hitting a bottleneck is to build something, not escalate it
• You think about your tech stack second and your growth hypothesis first

CircleCI is an Equal Opportunity Employer. We celebrate diversity and are committed to creating an inclusive environment for all employees.