COMPEL Specialization Stream · AITM-ECI
COMPEL Academy — AI Experimentation Associate
Professional certification certifying that the holder can design, run, and govern AI experiments, with disciplined continuous-improvement loops across the COMPEL Evaluate and Learn stages.
Profession title: AI Experimentation Associate
Audience: ML engineers, product analysts, and transformation practitioners running controlled AI experiments.
Enroll in the AITM-ECI track
Registration, enablement, and the proctored assessment are delivered through compel.one. Seats open continuously.
Prerequisite chain
- AITF AI Transformation Foundations
- → AITM-ECI (this credential)
Learning outcomes
The learning journey is sequenced to cover each outcome below in order. Every article in the journey maps to at least one outcome.
- 1. Design a governed AI experiment with clear success criteria.
- 2. Instrument evaluation protocols that withstand audit.
- 3. Run continuous-improvement loops that feed the COMPEL Learn stage.
Body of Knowledge articles (18)
Module M1.3 (18 items)
- Article What an AI Experiment IsM1.3-Art01
- Article Hypothesis Formulation and Metric DesignM1.3-Art02
- Article Offline EvaluationM1.3-Art03
- Article Online EvaluationM1.3-Art04
- Article Hyperparameter Search and Model SelectionM1.3-Art05
- Article Experiment Tracking, Reproducibility, and ReplicabilityM1.3-Art06
- Article Pipelines and OrchestrationM1.3-Art07
- Article Continuous Integration for MLM1.3-Art08
- Article Continuous Delivery and Governed PromotionM1.3-Art09
- Article Evaluating LLMsM1.3-Art10
- Article Red-Team Experimentation for SafetyM1.3-Art11
- Article Experiment Cost and Compute BudgetM1.3-Art12
- Article Regulatory Documentation for ExperimentsM1.3-Art13
- Article Experiment Brief and Experiment ReportM1.3-Art14
- Lab Lab 01: Design and Execute an Offline Evaluation HarnessM1.3-Art51
- Lab Lab 02: Design an Online A/B Test with Sample-Size Calculation and Rollback CriteriaM1.3-Art52
- Case Study Case Study: Zillow Offers and the Missing Shadow EvaluationM1.3-Art61
- Template Artifact Template: Experiment BriefM1.3-Art71
Competencies demonstrated
- → Experiment design and hypothesis management
- → Evaluation protocols and holdout strategies
- → Continuous-improvement loops and learning backlogs
- → Measurement discipline for AI outcomes
Exam blueprint summary
- Assessment
- Non-proctored assessment
- Passing score
- 70% passing score
- Portfolio
- Not required
- Renewal
- Every 24 months
- Recommended hours
- 20
- CE credits
- 20
Linked Core Mastery context
The Specialization Stream assumes AITF Foundations fluency. These Core Mastery resources are the recommended grounding before entering the AITM-ECI learning journey.
Formal credential definition
The machine-readable Open Badges 3.0 / W3C Verifiable Credential
definition for AITM-ECI is published at
/credential/aitm-experimentation-ci
. HR platforms and AI citation engines can fetch the JSON-LD
document at
/credential/aitm-experimentation-ci.json
.