A small Seoul academy that prefers written critique over confident slides.
FlowMint Academy started in 2019 with twelve learners in a Gangnam co-working room. We have stayed deliberately small ever since because mentor review — written, specific, sometimes uncomfortable — is the product, and mentor review does not scale by adding seats.
From twelve learners to nine cohort courses, without breaking the model.
The first cohort was three weeks long, ran out of one rented projector, and produced two graduates who are still on our mentor team. We expanded by building courses we wished had existed when we were learning, not by chasing a market. The Bot Developer Track came out of a 2020 conversation with an alumnus stuck in their first week of unattended bot work. The Architect Capstone arrived four years later, when those same alumni asked us how to operate at the program level.
We deliberately do not run a sales team. Every cohort fills through the alumni network, search, and word of mouth — and we cap each cohort at fifteen learners because that is the upper limit at which one mentor can give written critique that is worth the paper.
Four sentences we agree on every quarter.
Project-first, theory-second
Every course week ends with something runnable. Theory exists to defend a design choice, not to fill an evening.
Mentor feedback that hurts a little
Generic praise does nothing. Our mentor notes are specific, written, and sometimes uncomfortable — that is the job.
Honest about what we do not teach
We name our limitations on every course page. If something is out of scope we say so before you sign up, not after.
Re-usable artifacts over slides
Templates, rubrics, and reference notebooks travel back into your work. Slides do not.
Slow growth, on purpose.
First cohort in a Seoul co-working room
Twelve learners, one mentor, three weeks. Half of them are still in the alumni network and two are now instructors here.
Bot Developer Track open-sourced
We released our REFramework grading rubric and 30+ teams adopted it for internal upskilling. The peer-review feedback rewrote three of our courses.
Enterprise team training launched
Eight enterprise client teams ran private cohorts with us last year. The learnings fed directly into Orchestrator Deployment Pro and the Capstone.
Seven people, one program, one director.
Below is our actual reporting chart. We deliberately keep it shallow — most decisions are written down in a shared document, not escalated. Everyone on the team teaches, including the director.
Former RPA Center of Excellence lead at a KR consumer electronics group; sets curriculum direction and mentor standards.
Eight years building attended bots for KR retail and logistics teams. Owns the foundations courses end to end.
Specialises in OCR, classification, and validation tooling. Designed the Document Understanding corpus.
Five years on enterprise client automation portfolios in APAC. Owns assignment design and rubric calibration.
Former platform engineer for an unattended fleet of ~140 robots. Now leads our team training engagements.
Ten years across UiPath, Automation Anywhere, and reconciliation tooling. Runs the developer track clinics.
Career-shift specialist who works with cohort graduates on portfolios, interview practice, and offer negotiation.
Triages mentor requests, schedules clinics, and shepherds learners through the cohort calendar.
Pick a course, or come talk to us about a private cohort.
No estimator. No quote shopper. Just a 25-minute conversation about your team and what you are trying to build.