Shared Collections

Shared Collections gave US institutions, particularly Disability Support Services teams, a privacy-safe way to distribute lecture notes from note-takers to the students who relied on them.

  • Main quest
Type
Feature
Project
Genio Notes, then Glean
Role
UX Designer, leading problem framing, scenarios, roles and permissions, prototyping, UX writing and UI through delivery
Area
DSS note sharing, accessibility, admin workflows, privacy and permissions
Status
Launched January 2023 and still in active use

Snapshot

Problem
Students supported by Disability Support Services in the US often rely on a peer note-taker, but the existing sharing model was one-to-one, did not scale to a class and exposed the identities of people sharing and receiving notes.
Users
US institution admins and DSS teams distributing notes, note-takers and lecturers contributing recordings and notes, and recipients who needed access without having their support relationship exposed.
Goal
Design a multi-recipient sharing model that institutions could manage centrally, protected identities by default and made ownership, editing and access clear.
Outcome
Launched as a Glean for Education feature in January 2023 with admin-created collections, note-taker and recipient roles, anonymous sharing and a copy-to-your-account model that remains in active use today.

Short on time? Skip to the ending

Questions driving the project

What I needed to understand

  • How do students currently receive notes from a peer note-taker, and where does that break down?
  • Where does sharing expose identity or privacy that it should not?
  • What does an institution admin need in order to manage sharing at scale?
  • What should happen when a shared note is later edited or removed?
  • Who owns a shared note once a recipient has access to it?

What I needed to validate

  • Would an admin-managed anonymous collection serve the peer note-taker use case better than extending one-to-one sharing?
  • Which roles and permissions were genuinely needed, and which would be over-engineering?
  • Should shared events stay in sync with the original, or be copied as a point-in-time snapshot?
Glean admin screen for creating a shared collection with fields for collection name, note-takers, recipients and admin notes.

Shared Collections creation flow inside Glean Admin, showing how admins set up note-takers, recipients and context for the collection.

Challenge

Swipe, use the arrows, or press left and right.

Challenge card 1 of 4: 01 / Business trigger. One-to-one sharing proved demand, but US DSS teams needed to distribute notes to many recipients anonymously.

01

01 / Business trigger

One-to-one sharing proved demand, but US DSS teams needed to distribute notes to many recipients anonymously.

Sharing notes sounds simple until you ask who sees whom.

In the US Disability Support Services context, the peer note-taker relationship is sensitive: a student receiving support may not want that known, and a note-taker may not want to be identified either.

At the same time, an institution paying for the service needs to manage who is in a collection and check that the notes being shared are good enough to learn from.

The work could not be a bigger version of send a copy.

It needed a role-based model where admins could manage the collection, note-takers could contribute, recipients could access notes anonymously and copied notes became stable artefacts in the recipient account.

Research

Methods

  • Framed the problem from the existing send-a-copy pattern and the US DSS peer note-taker use case
  • Mapped sharing scenarios across admins, note-takers and recipients
  • Worked with Product and Engineering to define requirements and edge cases
  • Created wireframes and an interactive prototype to test the roles model and admin creation flow
  • Compared lighter-weight sharing options against a managed collection model

Key insights

  • Privacy was the product. The use case only worked if neither side was exposed to the other.
  • Institutions needed to own and manage the collection, while students needed the receiving experience to feel safe and simple.
  • Clear roles reduced risk: admins manage, note-takers contribute and recipients consume.
  • Recipients needed stable notes they could rely on for revision, even if the original changed or was removed.
  • Admins needed visibility for quality assurance without exposing peers to each other.

Hypotheses

  • If sharing is managed by admins, institutions will have the control they need without relying on one-to-one workarounds.
  • If recipients and note-takers are anonymous to each other, the support relationship will be protected by default.
  • If recipients copy notes into their own account, they will have a stable revision artefact they can keep and control.
  • If the roles are minimal and explicit, the model will be easier to understand and maintain.

Product direction

  • Make privacy the foundation of the sharing model.
  • Use admin-managed collections rather than expanding one-to-one sharing.
  • Define minimal roles with clear permissions.
  • Give recipients a stable copy rather than a live link.
  • Let admins assure quality without exposing identities between peers.
  • Use UX writing to explain ownership, visibility and action states clearly.

Core flows

Roles and permissions

The core design challenge was deciding who could create, contribute, view, copy, edit and keep shared notes.

  • Modelled three roles: admin, note-taker and recipient, each with distinct minimal permissions.
  • Made admins implicit members so institutions could assure quality without extra setup.
  • Handled edge cases around expired users, deleted users, cross-organisation access and recipients who only copy.

Anonymous sharing

The peer note-taker use case was only viable if students receiving support and people sharing notes were not exposed to each other.

  • Designed recipients to see the notes, not the person who contributed them.
  • Kept contributor and recipient identities hidden between peers by default.
  • Made the sharing relationship legible to admins while keeping it invisible where privacy mattered.

Copy to your account

Recipients needed something stable they could revise from, not a live link that could change beneath them.

  • Chose a point-in-time copy model rather than live syncing shared events.
  • Let recipients copy an event into their own account as an artefact they owned.
  • Designed clear copied states so recipients knew what they already had and when they might need to copy again.

Admin management

Institutions needed to create, manage and monitor shared collections at scale.

  • Designed the create-collection flow across name, note-taker selection, recipient selection and admin notes.
  • Created a collections table with edit and delete actions.
  • Wrote UX copy across admin, note-taker and recipient touchpoints, including banners and notification emails.

Iteration

What did not work

  • Treating sharing as send a copy to more people ignored identity, administration and ownership.
  • Leaving note ownership ambiguous made the recipient experience feel fragile.
  • A live-syncing model risked changing study material underneath recipients.

What changed

  • Reframed the feature around roles, anonymity and ownership rather than a larger share button.
  • Committed to a copy-not-sync model so recipients always had a stable artefact.
  • Gave admins quality-assurance visibility without exposing peers to each other.
  • Refined UX writing after launch as real institutions started using the feature.

Impact

  • Shipped a differentiating Glean for Education feature that solved a real accessibility need.
  • Created a durable roles-and-permissions model that later sharing work could build on.
  • Recent usage shows the feature is still active years after launch across hundreds of organisations.
  • In the last 12 months, note-takers added roughly 1,785 events and recipients opened shared collections roughly 2,183 times.

The strongest signal is endurance. This feature predates the current analytics baseline, so recent usage is included to show the model still works rather than to overclaim launch impact.

Outcome

  • Launched to Glean for Education in January 2023 as an online-first feature
  • Supported the American higher-education market, particularly Disability Support Services teams
  • Established an admin-managed collection model with note-taker, recipient and admin roles
  • Protected anonymity between note-takers and recipients by default
  • Created a copy-to-your-account model so recipients could keep stable notes for revision
  • Supported go-to-market with note-taker, recipient and admin explainer materials
  • Remains in steady multi-institution use with only incremental UI refinement since launch
  • Shared Collections made sensitive note sharing safer and more scalable for US institutions.
  • DSS teams could manage distribution centrally, note-takers could contribute without exposing themselves, and recipients could access and keep notes without revealing that they relied on support.

Reflection

  • The real design work was not the screens. It was the permissions model.
  • Copy, not sync, was the decision I am most confident in because recipients revising for exams need something stable they own.
  • Designing for a vulnerable use case made privacy a first principle rather than a setting.
  • A simple feature can hide a complex trust model underneath it.
  • The strongest signal is time: a feature designed in 2022 is still doing its job across hundreds of institutions in 2026.
Back to portfolio