Skip to main content
Hiring proof · Experimentation, CRO & analytics

I build the system that decides which growth ideas deserve investment.

Program leadership, conversion rate optimization, causal measurement, and growth analytics connected to decisions the business can defend.

Evidence in one sentenceAt NRG, that system scaled annual experiment velocity from roughly 20 to more than 100 tests; internal 2025 program reporting recorded $30M+ in impact.

The operating surface

One journey. One accountable system.

  1. 01Question
  2. 02Feasibility
  3. 03Experiment
  4. 04Decision
  5. 05Learning system
Selected evidence

What I owned, what changed, and how I know.

01
Enterprise program · NRG Energy

The work was not running tests. It was making the tests trustworthy.

Five regulated brands needed one way to turn ideas into comparable decisions. I built shared intake, prioritization, statistical rules, QA, readouts, and executive reporting so teams could move faster without quietly lowering the evidence bar.

Inspect the NRG program
02
Causal measurement · Silicon Valley Bank

I connected an offline campaign to web behavior, CRM evidence, and a market comparison.

For Startup Banking, I designed a comparative geo test, instrumented QR-enabled offline-to-online paths, and translated the result into an investment decision. The same role connected acquisition, product usage, CRM activity, and funded-account outcomes across a $1B+ pipeline.

Review the SVB measurement work
03
Founder build · GrowthLayer

I am building the product I wanted when experiment memory lived in slide decks.

GrowthLayer turns test plans, results, decision records, and reusable evidence into a searchable operating system. Building it forces the methodology to become product behavior: required identifiers, fail-closed decisions, analysis-only records, and reproducible planning.

Open GrowthLayer
How I operate

The repeatable moves behind the outcomes.

  1. 01

    Start with the decision

    Name what changes if the result is positive, negative, or inconclusive.

  2. 02

    Test feasibility early

    Size traffic, baseline, detectable effect, allocation, and decision horizon before creative work.

  3. 03

    Protect the readout

    Define exposure, primary metric, guardrails, QA, and one prespecified decision method.

  4. 04

    Make learning reusable

    Record identifiers, limitations, business context, and the next decision—not only a winner.

AI-native execution

The proof is in the systems built.

AI increases delivery leverage. It does not replace the evidence standard.

01

Methodology expressed as software

I use coding agents to turn experiment policy into typed workflows, validation, tests, and decision records.

GrowthLayer decision workbench, calculators, and guarded ingestion
02

Faster analysis with a human evidence owner

Models help inspect patterns, draft readouts, and challenge assumptions; the metric definition and causal claim remain human decisions.

NRG analysis workflow and executive reporting system
03

Evaluation before automation

AI workflows get test cases, structured rubrics, failure handling, and explicit boundaries before they earn autonomy.

Production publishing and product workflows with fail-closed gates
Role alignment

Where this evidence is most useful.

The strongest fit is a role where experiment design, causal measurement, product judgment, and organizational adoption matter together. This page does not claim production ML-infrastructure or native SDK engineering depth.

  • Head or Director of Experimentation
  • Experimentation Program Lead
  • Head of CRO / Digital Optimization
  • Growth Analytics or Marketing Analytics Lead
  • Experimentation-focused Product or Data Science leader
Hiring conversation

Need someone to own this system—not another channel report?

Share the remit, decision rights, and outcome you need this role to own.

Discuss role alignment