Data & AI 90 days 2-3 hours/day updated 2026-06-01

ModelOps 90-Day Learning Path

Learn ModelOps in 90 days: enterprise-scale model governance, cross-team model registries, risk scoring for AI models, model lifecycle management, and regulatory compliance for AI systems.

What ModelOps means

ModelOps is the operational and governance layer for all AI and ML models across an enterprise, regardless of who built them. While MLOps focuses on the engineering pipeline for a single team, ModelOps provides the cross-functional governance framework covering model inventory, risk assessment, explainability requirements, regulatory compliance (EU AI Act), access controls, and centralized lifecycle management for all models deployed in the organization.

Who should follow this path

AI governance leads and Chief AI Officers
ML engineers in regulated industries (finance, healthcare)
Platform engineers building enterprise AI platforms
Risk and compliance officers overseeing AI systems
Data scientists at organizations with many production models

Prerequisites

Experience with ML model training and deployment
Understanding of model monitoring concepts
Familiarity with enterprise governance frameworks
Basic cloud platform knowledge
Exposure to regulatory environments (financial services or healthcare preferred)

The 90-day plan

Daily study recommendation: 2-3 hours/day, six days a week. Consistency beats intensity — block the time in your calendar like a meeting.

Days 1–15: Foundation

ModelOps vs MLOps distinction and scope
Enterprise AI model inventory and cataloging
AI/ML model risk tiers and classification
EU AI Act overview and compliance implications
Model governance frameworks and charters

Outcome: Build an enterprise model inventory with risk-tiered classification aligned to the EU AI Act.

Days 16–30: Core concepts

Centralized model registries: MLflow, Domino Data Lab, SAS Model Manager
Model metadata standards and documentation requirements
Model cards and factsheets
Approval and validation workflows for high-risk models
Model lineage and audit trail management

Outcome: Configure a centralized model registry with documented model cards and approval workflows.

Days 31–45: Tools and workflows

Model performance monitoring at enterprise scale
Drift detection across a portfolio of models
Champion-challenger and model comparison frameworks
Feedback loop management and ground truth collection
Model deprecation and retirement processes

Outcome: Implement centralized drift monitoring and champion-challenger testing for a multi-model portfolio.

Days 46–60: Hands-on projects

Explainability and interpretability requirements (SHAP, LIME)
Fairness and bias detection tools (Fairlearn, AI Fairness 360)
Model explainability for regulatory requirements
Human-in-the-loop workflows for high-risk decisions
Adverse action notifications in credit/lending models

Outcome: Audit a production model for bias and fairness, and generate an explainability report for compliance.

Days 61–75: Advanced practices

ModelOps platform selection: DataRobot, H2O.ai, Domino
MLOps platform integration and unified governance
Regulatory model validation (SR 11-7 for financial services)
Cross-team model sharing and reuse governance
AI incident response and model rollback procedures

Outcome: Design a ModelOps platform architecture meeting SR 11-7 or EU AI Act validation requirements.

Days 76–90: Portfolio, interview & certification prep

ModelOps portfolio: enterprise governance project
Preparing for DataRobot or SAS certifications
ModelOps interview questions for AI governance roles
Metrics: model inventory coverage, validation cycle time, drift response SLA
Emerging: foundation model governance, AI safety frameworks

Outcome: Deliver an enterprise ModelOps governance project and be interview-ready for AI platform roles.

Weekly outcomes at a glance

Phase	Outcome
Days 1–15	Build an enterprise model inventory with risk-tiered classification aligned to the EU AI Act.
Days 16–30	Configure a centralized model registry with documented model cards and approval workflows.
Days 31–45	Implement centralized drift monitoring and champion-challenger testing for a multi-model portfolio.
Days 46–60	Audit a production model for bias and fairness, and generate an explainability report for compliance.
Days 61–75	Design a ModelOps platform architecture meeting SR 11-7 or EU AI Act validation requirements.
Days 76–90	Deliver an enterprise ModelOps governance project and be interview-ready for AI platform roles.

Tools to learn

MLflow
DataRobot
Domino Data Lab
SAS Model Manager
H2O.ai
Fiddler AI
Evidently AI
Fairlearn
AI Fairness 360
SHAP
Weights & Biases
Grafana

Labs to practice

Mini projects

Build an enterprise model registry with automated model cards, risk tiers, and approval workflow using MLflow
Implement a bias and fairness audit pipeline using Fairlearn and AI Fairness 360 for a credit scoring model
Create a champion-challenger model testing framework with automated promotion/demotion based on business KPIs

Interview questions to prepare

What is the difference between ModelOps and MLOps?
How do you build a model risk tiering framework for an enterprise?
What is a model card and what information should it contain?
How does the EU AI Act affect your model governance processes?
Explain SR 11-7 guidance and its impact on model validation in financial services.
How do you handle model deprecation without disrupting downstream consumers?
What is champion-challenger testing and when do you use it?
How do you detect and mitigate bias in a production classification model?

Certification suggestions

DataRobot Certified AI and ML Practitioner — DataRobot
AWS Certified Machine Learning Specialty — AWS
Databricks Certified Machine Learning Professional — Databricks
SAS Certified AI and Machine Learning Professional — SAS

Browse the full certification registry for exam details and official links.

Free resources

// instructor-led option

Prefer live, guided training with mentors and certification support? DevOpsSchool.com runs paid instructor-led programs that pair well with this free path.

Explore paid training on DevOpsSchool.com ↗