I learned how to learn from a sound art course. Not because the curriculum was perfect — because the teacher was so genuinely alive in their subject that I caught it. A door opened. I got curious, started making things. That's when I understood: learning isn't transferred. It's ignited.
It is literally impossible to build memories or make meaningful decisions without emotion.
Mary Helen Immordino-Yang put it in neurobiological terms. Most AI and Data Science courses for non-technical learners commit the exact error her research predicts will fail — they dump content without establishing why it matters to this person, in this life, right now. I've spent a decade watching people sit through polished modules, complete them, and retain almost nothing. Not because they weren't capable. Because nothing was asked of their intelligence.
Rancière has a name for this. He calls it stultification — and argues that the more carefully we explain, scaffold, and simplify, the more we communicate: you cannot do this without me. I work from the opposite assumption: the learner is already intelligent. My job is to design conditions where that becomes visible to them.
Art pedagogy, somatic practice, Gestalt methodology, 10+ years building Data Science and AI/ML curricula for B2C platforms and B2G clients. The arts taught me that anyone who genuinely wants to learn something, can. My job is designing the conditions where that becomes true.
I use ChatGPT, Claude, and NotebookLM daily — to move faster on structure so my attention stays where it matters: on why a learner would care, and whether the material makes that visible.
I'm training as a somatic therapist — working with how the body holds attention, stress, and memory. Two hours of passive content and your nervous system shuts down. Learning that connects to real stakes, real people, real problems — learns differently. That's what I design for.
Rodchenko Art School — 15 professional programs + 10 short courses
Cross-institutional publishing + VR course for artists
Nosorog Publishing · Rodchenko · Sreda Obuchenia — 1000+ students
Graphic Design — industry case model, partner briefs
Interior Design Specialization — project-based, real commissions
Case-based model became repeatable pattern across B2C professional tracks
MIPT Master's — Product Management program
RANEPA · SPbPU · Innopolis Game Design
Product & Project Management · Game Design — across technical & creative disciplines
AI literacy for creatives, business practitioners, developers
Same technical content, three completely different learner contexts — visual, operational, engineering
Data Analyst + ML Engineer specialization — active redesign
IT Product Management Master's · Research University · Skillbox
A prestigious research university's Product Management master's program was underperforming. Students weren't completing assignments. Mentors were frustrated. Satisfaction scores were critically low — CSAT sat at 53. I joined as Learning Experience Designer with no prior background in product management. What I brought instead was structural thinking: how learning is supposed to be built, and where breakdowns happen between intention and outcome.
The program was built around real outcomes — students were expected to graduate with working startup prototypes, not just grades. The stakes were high on both sides.
Assignments were methodologically broken. Tasks didn't match the skill level being taught — students hit walls, mentors hit walls trying to explain them.
Mentor onboarding didn't exist. Mentors were thrown into grading with no shared rubric, no clarity on expectations. Every cohort started from scratch.
The cost model was invisible. Mentors were being paid for iterations that shouldn't have been necessary. Assignments created confusion loops, and every loop cost money.
Assignments rebuilt from scratch — clear success criteria, examples, scope limits. The confusion → resubmission → re-review cycle simply stopped.
Mentor onboarding redesigned — shared rubrics, grading guides, escalation paths. Mentors stopped improvising.
Financial audit + fix — analyzed budget spreadsheets line by line, identified duplicate review payments, restructured workflow so we paid once for work done once.
I wasn't just a methodologist on this project. I went into the budget tables, found where money was leaking, and redesigned the assignments so the leak stopped.
Before → After
The reason this worked: I didn't stay in my lane. Most learning designers fix the content. I fixed the content, the mentor system, and the financial structure — because all three were part of the same broken loop. Coming in as a domain outsider turned out to be an advantage: no assumptions about how product management "should" be taught, so the structural problems were visible.
Data Science curriculum redesign · Skillbox · Ongoing
Data Science courses age fast. The technical content stays correct — but the way it's written stops working. Examples go stale, explanations assume context students don't have, and somewhere along the way the module stops feeling like it's talking to a person and starts feeling like a textbook nobody asked for.
The problem wasn't the information — it was the relationship between the information and the learner. Modules answered "what" and "how" but skipped "so what." And when you skip "so what," you lose people in the first paragraph. One pattern kept appearing in feedback: students weren't asking technical questions. They were asking "when would I actually use this at work?" The module had answered the wrong question for months. Nobody had noticed because nobody had looked at the feedback as a pattern — only as individual complaints.
Listen before fixing. Student feedback through AI-assisted analysis first. Patterns surface fast: which explanations consistently confuse, which sections generate the same question over and over, where people quietly give up.
Diagnose, then rewrite with context. The problematic module goes to AI with a specific brief: show me where a non-technical learner loses the thread. The output is a diagnosis, not a rewrite. Then I work — adding the business context the module was missing.
Give experts a draft, not a blank page. The revised module goes to the SME with specific questions. They check a draft that's already 80% right — their time goes to precision, not production.
What one SME said
"He hadn't realized how much time he'd been spending explaining the same thing in Slack to students who'd already read the material. The questions stopped — not because the topic got easier, but because the module finally answered them before they had to ask."
On the numbers
No before/after metrics yet — redesigned modules still rolling out. Previous versions had satisfaction scores around 50. Target for new versions: 70–75. Based on structural changes made, that's realistic. I'll update when data comes in.
The principle
I don't use AI to generate courses. I use it to see what I can't see alone — patterns in feedback, friction in text — and to free up the time I need to do the work that actually requires a human: understanding why a student would care, and making sure that answer is in the material before they have to ask.
Sound Art · Online course + 1:1 mentorship · Independent practice
Most people think they can't work with sound. They imagine expensive equipment, technical expertise, years of training. The real barrier is different: they've never been taught to actually listen. This course started from that assumption — and spent ten weeks proving it wrong.
Cameras off — deliberately
An online session with twelve cameras on is a performance space. Nobody listens in a performance space — they manage how they look while listening. Visual attention and auditory attention compete. Remove one, and the other sharpens.
Collective feedback as first act of creation
A student brings a sound fragment. Ten people open a shared Google Doc and write simultaneously: associations, images, questions. Nobody is critiquing. Everyone is responding. The document fills in real time and the student's work stops being a private experiment — it becomes a thing that exists in the world.
Simplest possible entry point
No expensive equipment. No software learning curve. Here is how you begin working with sound using what you already have. The instructions were short because the territory was large — I wanted students moving through it, not studying the map.
Expanded perception is a legitimate learning outcome. It's also one of the harder ones to engineer — because it requires the designer to trust the learner's intelligence rather than manage their encounter with difficulty.
The 1:1 version
A year of working with one person who came with no background in sound and an associative, non-linear way of thinking. Each session I tracked which ideas were connecting to which — not to direct, but to reflect back the pattern that was already forming. What looked like tangents turned out to be the actual material.
By the end: a wider perceptual range. She went to concerts and heard things she couldn't have heard before. Her listening changed — and with it, how she moved through her work.
SME System & Podcast Collaboration · Sreda Obuchenia
Expert content was dry and disconnected from learner reality. SMEs were burnt out. Webinar attendance: 1–2 students per session.
Collaborated with "Either/Or" (Libo-Libo) podcast to blend technical theory with personal storytelling. Designed structured onboarding for SMEs so they could teach without burning out.
100% Author NPS — across all SMEs.
Webinar attendance: 1–2 → 15–20 students/session.
US Market Research
Deep competitor research (Udemy, Domestika, Coursera) for a US-focused design course — validated product-market fit and pricing strategy.
EU Collaboration
Managed an exhibition project in Barcelona, navigating cultural differences and coordinating full production in English with international teams.
If you're an expert, practitioner, or educator who wants to turn your knowledge into a course — but doesn't know where to start, feels stuck in the structure, or just needs someone to think alongside — that's exactly where I work best.
I help authors shape raw expertise into learning experiences that actually land. That means: extracting what matters from what you know, building sequences that learners can follow, and giving feedback that moves things forward without flattening your voice.
Available for consulting, course co-development, and SME collaboration. One conversation is enough to figure out if we're a good fit.
A real idea, moved fast, iterated honestly — with everything we know about how humans actually learn brought to bear on making it land. Open to full-time roles and freelance projects in global EdTech — especially teams building AI literacy, Data Science, or tech-skills programs for people who don't think of themselves as technical.
molokova1@gmail.com linkedin.com/in/mariamolokova Montenegro · Remote-first · EN / RU