I lead growth, experimentation, analytics, and AI-enabled delivery. This page translates the résumé into what a hiring team or consulting buyer actually needs to know: what I owned, what changed, how I lead, and where the evidence stops.
I’m a growth experimentation leader based in Austin, Texas. I use behavioral economics to form stronger hypotheses, statistical discipline to protect the decision, and business context to connect the result to revenue. I also use AI and agent-assisted software delivery to research, specify, build, test, and operate products with explicit quality and safety controls.
The through-line is not a tool or channel. It is building decision systems that make learning faster and investment more defensible.
AI is not a side tool in my workflow. I use Claude, ChatGPT, Claude Code, Codex, and Claude Cowork to research, specify, build, debug, analyze, automate, and ship.
I keep a deliberate learning loop around new models and platforms, but the standard is working software and better decisions—not novelty.
Use Claude Code and Codex to build and maintain websites, applications, GrowthLayer, and production features—with review, testing, and deployment discipline around the agents.
Websites, applications, and founder-led products shippedDesign and implement experiments through code and platforms, including PostHog, Optimizely, Adobe Target, and VWO, with explicit metrics, QA, and decision rules.
100+ NRG experiments in 2025 · annual velocity scaled from 20 to 100+Build workflows with model APIs, structured outputs, tool use, context design, evaluations, guardrails, and failure handling—not isolated prompt demos.
OpenAI and Anthropic integrations · repeatable agent workflowsConnect event design, customer behavior, funnel performance, and experiment results to revenue, pipeline, and investment decisions.
$30M+ in NRG internal program reporting · $1B+ SVB pipeline supportedMove from customer research and product requirements to prototypes, vendor direction, launch, measurement, and the decision to scale, redesign, or stop.
Jobsolv, GrowthLayer, publishing systems, and web productsAlign marketing, product, analytics, UX, engineering, finance, executives, and agencies around a shared evidence standard and a clear next move.
Enterprise programs, distributed teams, and capability buildingBuilt the operating system for a multi-brand enterprise experimentation program: intake, prioritization, statistical rules, QA, and revenue reporting. Scaled annual velocity from roughly 20 to more than 100 tests.
Connected acquisition, product usage, CRM, and funded-account outcomes. Built measurement and controlled geo-testing approaches for growth investment decisions.
Ran creative, audience, and landing-page experiments across performance accounts and trained paid-media specialists in behavioral-economics-led testing.
Built acquisition, positioning, KPI, competitive research, and go-to-market strategies for early-stage startups.
Used purchasing and contract data to identify savings opportunities, recommend supplier-term improvements, and support purchasing compliance.
Led growth and business development for an e-commerce startup across funnel testing, Shopify, paid acquisition, UX, SEO, and remote-team leadership.
These projects sit beside the employment timeline. Some are founder-owned products; others are focused product, positioning, and conversion engagements.
The direct extension of my enterprise experimentation work into a product I own.
Builds an experimentation and CRO analytics platform that turns scattered A/B test results into searchable institutional knowledge and behavioral-science insights for growth teams.
Products where I own the roadmap, delivery constraints, and operating tradeoffs.
Built an AI-powered career platform to a July 2026 first-party snapshot of 35K+ users and $0 current paid user-acquisition spend after earlier paid tests stopped.
Builds and validates a self-use-first content operating system for research synthesis, evidence-backed publishing, and cross-channel distribution workflows.
Focused roles for other founders, separated from companies I own.
Helped reposition a multi-feature product into a focused higher-value offer, owning conversion, messaging, and performance through the transition.
Provided marketing consulting for an AI-enabled real-estate analytics platform.
Led experimentation across marketing, product, UX, analytics, engineering, and agency partners for five national brands. Built the shared decision system that turned competing opinions into evidence-backed investment choices.
Directed a 27-person distributed roster across the build, with up to 22 active in a week across engineering, UX/UI, and growth. Translated product vision into requirements, sprint plans, operating rhythms, design standards, and delivery milestones.
Built and managed a distributed bench of freelancers and agencies using portfolio evidence, paid work trials, clear contracts, and measurable deliverables to reduce hiring theater.
Trained marketers in behavioral economics, coached paid-media specialists, developed early-career interns, and mentored practitioners through ADPList.
“When challenges arise and opportunities emerge, you want him in the room.”
Formal training across experimentation, behavioral science, analytics, and data storytelling—paired with years of applying those frameworks under business constraints.
University of North Carolina Wilmington · Omicron Delta Epsilon economics honor society.
Native proficiency in both; currently learning Spanish and Japanese.
Free 1:1 guidance for people moving into experimentation, CRO, and analytics.
View the mentoring profile (opens in a new tab) →Second-place finalist and public-favorite concepts built in 54-hour sprints.
I read for models I can test—and for the assumptions, incentives, limitations, and replication questions behind them.
How people value options, respond to framing, and make choices.
Separating signal from noise and updating beliefs under uncertainty.
Customer discovery, operating systems, founder judgment, and difficult teams.
The economic and narrative systems shaping what people believe and buy.
The strongest working relationships usually start with context: where someone came from, what they notice, and what they care enough to keep learning.
I have lived in Kunming, Shanghai, San Diego, New Bern, Wilmington, and Austin. I am native in English and Mandarin and currently learning Spanish and Japanese.
Fourteen countries and 43 U.S. states so far—including a vanlife chapter across the United States and Canada in a van named Howl.
I share Austin life with a Texas Heeler rescue from Austin Pets Alive—equal parts Australian Shepherd intensity and Cattle Dog negotiation.
I keep several books in progress across behavioral science, business, history, technology, and fiction, with a running log of what I read and learn.
AtticusPoet began as a compassionate resource for people navigating relationship OCD and retroactive jealousy—a reminder that useful products start with a real person’s problem.
I have a strange habit of finding favorite restaurants shortly before they earn their first Michelin star. Recommendations come with a calendar reminder.
Articles on this site are based on Atticus Li’s experience, research, experiments, and interpretation. AI assists with research synthesis, organization and drafting.
Employer-sensitive experiment data is anonymized and expressed as safe ranges. Full-AI generation is not presented as Atticus’s first-hand work.