← All 186 courses
AI-220 AI & Data

Feature Stores & ML Platform Engineering

Make features a shared, correct, discoverable asset instead of code copied between a training notebook and a serving handler. Learners stand up offline and online stores, reproduce a historic training set with point-in-time correct as-of joins, and catch the leakage a naive join introduces. The course ends on the decision most teams get wrong: whether a feature store is warranted at all, argued from reuse counts, freshness needs and operating cost.

B4Professional
5modules
35skill atoms
2role journeys
Curriculum

What this course covers.

5 modules, 35 named skill atoms. Expand any module to see them.

1The Problem a Feature Store Solves7 skill atoms
duplicated feature logic across teamstraining serving skew symptomspipeline jungle of one-off jobsoffline batch vs online lookup pathsFeast Tecton and Databricks optionsbuild vs buy vs neithercriteria for deciding against one
2Offline and Online Stores7 skill atoms
Parquet and warehouse offline storageRedis and DynamoDB online storesmaterialisation jobs and push ingestionsingle-digit millisecond lookup budgetentity keys and join semanticsstreaming features from Kafkastorage cost against freshness gain
3Point-in-Time Correctness7 skill atoms
label timestamp vs feature timestampas-of joins on event timefuture leakage from a naive joinTTL and window semanticslate-arriving and out-of-order eventssplitting by time rather than at randomreproducing a historic training set exactly
4Feature Definitions as Code7 skill atoms
definitions in version controlentity and feature view schemastransformation logic shared across both pathsCI validation of definitionsbackfill jobs and idempotencyschema evolution and deprecation windowslineage from source table to model
5Freshness, Reuse and Discovery7 skill atoms
freshness SLA per feature groupstaleness alerting and monitoringfeature drift detection against trainingsearchable feature cataloguenamed owner and on-call per featurereuse counted across modelsdeprecating a feature without breaking consumers
Where it fits

AI-220 in the role journeys.

This course appears in 2 of our 45 role journeys. Here is what a learner takes immediately before and after it in each.

MLOps / LLMOps Engineer

Professional stage
AI-211AI-220AI-403

AI Infrastructure Engineer

Professional stage
AI-219AI-220AI-209

Roles this course serves

The capability ladder

This course is authored to band B4.

Every course we run is written to one rung of the CASI ladder, so a plan can be assembled to take a team from where they are to where they need to be.

What do B1–B6 mean?The CASI Capability Ladder — click to expand

Every course targets a band on the CASI Capability Ladder — our six-band proficiency scale, anchored to open standards (O*NET, ESCO, NICE, NIST AI RMF, Bloom's). A band tells you how deep a course goes, and what evidence proves it.

What the learner can doTypical evidence
B1
AwareUnderstands concepts and vocabulary; uses tools with guidance
Knowledge checks
B2
FoundationPerforms standard tasks correctly in familiar contexts
Guided labs, autograded exercises
B3
PractitionerDelivers complete pieces of work independently
Scenario labs, proctored hands-on exams
B4
ProfessionalHandles production-grade complexity, trade-offs and failure modes
Break-fix drills, design defenses
B5
AdvancedEngineers systems end-to-end under constraints; leads others
Rubric-scored capstones, vivas
B6
ExpertSets direction; recognised authority across teams
Portfolio + panel evaluation

A note on B6. Courses in this catalog target B1–B5. B6 is not taught — it is recognised, through a portfolio and a panel, once someone is setting direction for others. Every journey here is built to land a learner at B5.

Next step

Run AI-220 for your team.

This course runs at several lengths depending on how deep you need to go and how much of it your people already have. Tell us who is being trained and we will scope it.

Add it to a training plan Talk to our team Check your team’s level free