Signals from recent cohorts

Mixed metrics—some from satisfaction surveys, some from delivery logs—so you can see how the studio operates week to week.

12
Open catalog modules spanning JVM, Spring, and data access
38%
Learners joining from non-Tokyo prefectures in the last intake survey
9.1
Average mentor note helpfulness (10-point cohort survey scale)
140+
Recorded micro-demos across all tracks (minutes vary by topic)
6
Partner organizations co-designing at least one lab per year

Inclusions you can verify

Transparent scope: repositories, rubrics, and office hours—without promising outcomes we cannot measure.

Reviewable artifacts

Every sprint ends with something linkable: PRs, ADR drafts, or profiling notes you can attach to a portfolio.

Mentor annotations

Feedback targets your diff, not generic slides—timed so it lands before the next assignment opens.

Paths, not vibes

Sequences spell prerequisites, expected hours, and what we deliberately skip so you can plan around gaps.

1,180

Pull requests reviewed across Surge Fluxon cohorts since 2019—each tied to a rubric, not anonymous clicks.

JVM depth with mentor loops

Build Java services with structured feedback, cohort accountability, and self-paced options when midnight coding fits better than live labs.

Browse catalog Trusted by internal upskilling pods at logistics and fintech teams

The flow — four beats

Horizontal rhythm instead of numbered circles; each beat has a deliverable.

01

Orientation

Tooling, repo access, and honesty about prerequisites.

02

Build slices

Small features with rubrics referencing hiring panels.

03

Review window

Mentor passes land before the next slice unlocks.

04

Retro & story

Narrate trade-offs for interviews and internal RFCs.

Featured modules

A rotating slice of the catalog—filters on the full listing page go deeper.

Full catalog →

From our cohorts

Varied lengths and formats—experience-first, tied to specific modules on this site.

Concurrency lab’s structured task scope exercise made shutdown behavior visible; I borrowed the worksheet for our on-call primer.
Mika · Kyoto
Gradle Enterprise Builds: scan reading—not glamorous, but it funded our cache node.
Ren
OCA study studio’s timed mode exposed that I hemorrhaged minutes on generics; the whiteboard clinic fixed that faster than rereading guides. Still wish one mock exam had even hairier wildcard questions.
Paula Nguyen · 4/5
Bootcamp on-call simulation felt close to our incident retro format—useful stress, not theater. Client in enterprise integration.
Anonymous learner
Data Access with JDBC and jOOQ clarified when ORM magic hurt us; batch insert benchmark embarrassed an N+1 we defended for too long.
Ivan Petrov