introducing a knowledge library for everything you've figured out in your career.
ask copalm to build a library...
Attach
copalm.ai turns one expert's experience into many learners' experiences —
a social platform built on shareable knowledge graphs.
my role —
founder
product designer
user researcher
ux writer
vibe-coder
worked with —
professors
content designers
AI leadership
managers
developers
impact —
53+ sign-ups
23+ SMEs interviewed
117-component design system
100% bootstrapped with AI
timeline —
18 mos
january 2025
to present
started as capstone,
soon become a company
click
meet the Subject Matter Expert (SME) — they knew what to teach, but not where to begin.
top pain points
no starting point
knowledge is invisible.
tools missed step zero
the "blank page" problem is #1 among creators
who know their expertise
sourced from
900+ creator survey, 2022
competitive gap
14 platforms existed.
All of them assumed that the expert already knew where to begin.
success criteria
can I enable an expert to structure and publish knowledge with zero blank-page friction?
research showed 2 things
AI was generating, not thinking
people wanted a thinking partner, not a faster way to make the same thing.
sourced from + 23 interviews with
SMEs, Content Designers, Educators, and AI Leadership
so v1 tried to be exactly that.
at this point I realized, the design wasn't the problem — the target was.
the user wasn't wrong — targeting people who knew what to teach was.
they already had tools that worked;
the real opportunity was the expert who had none.
is there an unmet need for experts?
does course creation work for non-designers?
is there a bigger market than Instructors?
each step maps back to 1 of the 3 pain points we identified.
LAND
PERSONALIZE
PLAN
BUILD
ANNOTATE
PUBLISH
SME drops context — resume, file, link. copalm starts informed.
Social URLs teach copalm the SME's voice and tone.
Structured questions: audience, goals, format. No blank page.
Conversation on the left. Course structure forming live on the right.
Human + AI suggest instructional improvements together.
One click. Course goes live with a landing page.
context first —
the SME starts with a foundation, not nothing.
no starting point
the conversation
makes invisible knowledge tangible.
knowledge is invisible
the setup work every
other tool assumed you'd figured out yourself.
tools missed step zero
this time, v2 began with a conversation, not a blank page.
click
but honestly, v2 solved half the problem. v3 made knowledge dynamic.
in usability studies, experts finally got past the blank page —
but the result was a static course. the expert's knowledge didn't live on, and every learner got the same fixed path.
a course is a snapshot. knowledge is alive.
v2 solved it for the expert, once —
but learning shouldn't be static, and
one fixed path can't fit every learner.
so we asked —
why must the learning be static?
why one fixed path for everyone?
what if the expert's knowledge could be a living layer each learner shapes for themselves?
this time, v3 began with a conversation, a library and a visual layer of knowledge.
the story isn't finished — v4 is being built now.
What's next? one expert's knowledge, many learners' paths.
click
design usually stops at the handoff — mine shipped straight to production.
prompt —
component states + specs, defined in Figma.
generate —
Lovable turned them into production React + Tailwind.
validate —
reviewed, corrected, shipped. nothing went live unchecked.
last updated — 06/30/2026
