Case 03 · B2C · AI · Built · 2026 · 5 min read

Atélier: search your handwritten notes by meaning

Problem statement

Find handwritten notes by meaning, even when you forget the exact words.

Problem

The right note is hard to find

Students accumulate handwritten notes across classes and notebooks. Before a quiz, they need to retrieve an idea quickly without remembering where they wrote it.

Issue

Memory works differently from keywords

A student remembers the concept, not necessarily the exact phrase on the page. Finding it can mean scrolling through notebooks instead of reviewing the material.

Solution

Search handwriting by meaning

Atélier combines an iPad writing experience with semantic search and summaries grounded in the user’s own notes. Ask in your own words, find the relevant page and get the gist.

Progress Built and demoed. Search evaluation and a classmate beta are next.

Role: Product, user research, design and build (Swift)
Platform: iPad with Apple Pencil
Timeline: July to October 2026
Users: MBA and grad students
Skills: Discovery, AI product judgment, 0-to-1 execution

Problem

Students remember concepts, but keyword search needs exact words

Classmates at IIM Lucknow described three recurring frustrations before quizzes.

They remember concepts, not exact keywords.
Review means scrolling through notebooks.
Typing sacrifices the sketches and structure of handwriting.

Users

Before a quiz, students need the relevant page and its gist

The primary user's moment before a quiz, today and with Atélier.

Diagram guide

Read each row left to right. The top row is what happens today; the bottom row is the same moment with Atélier. The only step that changes is the second one: asking in her own words instead of guessing the exact word.

Segment	Job to be done	Call
MBA and grad students	"Find what the professor said about pricing in week 3, and give me the gist before the quiz."	Primary. Heavy iPad note-takers, high review pressure, and I can reach and test with them directly
Working professionals	Recall decisions from meeting notes	Later. Strong need, but notes live in many tools
Designers and visual thinkers	Sketch and annotate	Served by the core tools. They value the canvas more than search

Competition

My initial app comparisons revealed a concept-retrieval gap

Product	What it does well	Where it stops
Apple Notes	Free, built in, searches handwriting	Matches words, not meaning
GoodNotes	Best-in-class notebooks and handwriting search	Finding still depends on remembering the right word
Notability	Handwriting plus audio, searchable	Same keyword limit; review is still scrolling
Typed AI notes (e.g. Notion)	AI search and summaries	You have to type, which loses the handwritten thinking

When I tried the three handwriting apps in July 2026, search in each matched the words on the page, not their meaning. That is the open space Atélier goes after.

Strategy

Exam review gives semantic search a focused starting point

Start: graduate students preparing for quizzes.
Value: search and summaries grounded in their own notes.
Risk: competitors can add semantic search; the review experience must earn repeat use.

Hypotheses

Search quality and repeat use determine whether the product works

H1. Meaning beats keywords. Search by meaning puts the right note in the top 3 for most real student queries. Target: 16 of 20.
H2. Summaries earn a place in review. Students use a summary at least once in a week with a quiz.
H3. Writing feel is the price of entry. If the pen feels worse than their current app, they never stay long enough to try search.

Goals and metrics

Success means finding useful notes, not generating more text

North StarNotes found per weekly active userIt moves only when the core promise works: you asked, and you found it.

Searches per active userDo students reach for search at all?
Search success rateShare of searches where the user opens a result
Summaries used per weekIs the one-tap gist part of review?

Guardrail: share of summaries users rate as accurate. Health: week-4 retention.

Scope

I focused version one on handwriting, search and summaries

Feature	Value to students	Effort	Call
Pen and paper feel	Table stakes	Medium	v1, first
Search by meaning	High: the core promise	High	v1
Grounded summaries	High: review before quizzes	Medium	v1
Echoes of past notes	Medium	Low: reuses search on the page being written	v1
Chat assistant, auto-flashcards	Medium	High, medium	Later
Sharing and collaboration	Low for this user	High	Not planned

Echoes are not a third AI feature; they run the same search on the page you are writing. Two features are also two things I can evaluate honestly.

Solution

Atélier finds handwritten pages by meaning and summarizes them

Search your handwritingType what you remember, in your own words. Atélier finds the page even if you wrote it differently.

Echoes of past notesWhile you write, a quiet hint points to related notes from earlier classes.

Pen and paper feelA pen, eraser, colours, lasso and ruler that feel right before any AI shows up.

How a question becomes an answer.

Diagram guide

Read left to right. Your handwritten pages are understood by meaning, and that powers both search (top path) and summaries (bottom path). Everything sits inside the dashed boundary: outside knowledge is never used. Design is not proof, so the eval checks that summaries add no claims and miss no points.

Decisions and trade-offs

I prioritized writing quality and kept summaries grounded in notes

01Writing experience first

A solid pen, eraser, colour, lasso and ruler toolkit before any AI

Why: If writing feels bad, nobody stays long enough to need search. In hindsight I would have tested search on classmates' existing notes first.

Trade-off: Weeks of delay before the riskiest bet, search by meaning, could be tested

02Grounded summaries

Summaries draw only from the user's own notes

Why: Students need the summary to reflect their class, not the internet. The eval checks whether it does.

Trade-off: "Smarter" summaries that add outside knowledge

03Scope

Search and summaries only; no chat assistant or auto-flashcards in v1

Why: Fewer AI features, each one tested properly.

Trade-off: Feature-list appeal

Status and next

The app is built; search evaluation and a student beta come next

Measure	Target	Status
Feature-complete app, faculty demo	Ship v1	Done, October 2026
Right note in top 3 results	16 of 20 real queries	Next: run the eval
Summaries rated accurate	8 in 10 or better	Next
Week-4 retention	4 in 10 beta users	Next: 10 to 15 classmates

Next time, test semantic search on existing notes before building the writing toolkit.