Case study · 2026 · Present

Continuum

Bridging the gap between therapy sessions through calm, AI-assisted patient care.

Role

Lead Product Designer (End-to-End)

Timeline

2026 · Present

Scope

AI Integration, Design System, Information Architecture

Context

Side Project / Independent Build

Overview

From a patient problem to a connected therapy product

Continuum started from a patient problem: forgetting important things that happened between therapy sessions.

Research then showed that continuity was also a problem for therapists, who manage multiple patients, different tools, session context, scheduling, resources and administrative work.

The product evolved from a patient focused idea into a connected therapist and patient experience designed to make therapy more continuous.

Initial assumption: Patients need a better way to capture what happens between sessions.
What changed: Research showed that continuity was also a problem for therapists.
Product direction: Create a connected experience that supports the therapeutic process before, during and between sessions.

Research

Listening before designing anything

I wanted to understand what happens before, during and between therapy sessions, from both the therapist and the patient perspective.

Patient portal, built after the research phase

35+
conversations
10
psychologists
5
coaches
20
people with therapy experience

Research goal: Understand what happens before, during and between therapy sessions, from both the therapist and patient perspective.
Research methods: Conversations and interviews with practising therapists, coaches and people with therapy experience.
Participant groups: 10 psychologists, 5 coaches and 20 people with therapy experience.
Research limitations: To be added
Validation status: Research completed. MVP built. Formal usability testing still needed.

Key questions

How do therapists prepare for and follow up after sessions?
How do they remember patient context and evolution?
What happens between sessions?
Which parts of the workflow take the most time?
What information do they need to have at hand?
How do patients remember things they want to discuss in therapy?

Direction

How the problem changed

01
Initial assumption
Help patients capture what happens between sessions.
02
What I learned
Therapists were also dealing with fragmented tools, patient context, scheduling, resources and administrative work.
03
Product shift
The focus moved towards the therapist workflow while keeping the patient experience connected.

Insights

What the conversations surfaced

Patients can forget important things they wanted to bring to therapy.
Evidence
Reported in conversations with people with therapy experience.
Product implication
Give patients a way to capture a moment when it happens.
Therapists use different tools and methods to manage information about the same patient.
Evidence
Reported across therapist and coach conversations.
Product implication
Explore one workspace that keeps patient context together.
Some therapists were already using AI for different tasks.
Evidence
Reported in therapist conversations.
Product implication
Design AI as reviewable assistance inside the existing workflow.
Therapists need to manage scheduling and calendars.
Evidence
Reported in therapist conversations.
Product implication
Bring agenda and calendar into the therapist workspace.
Therapists spend time finding or creating personalised resources.
Evidence
Reported in therapist conversations.
Product implication
Use AI to draft personalised resources the therapist approves.

Product decisions

From insight to decision

What I learned
Therapists use multiple tools around the same patient.
Decision
Explore an integrated therapist workspace.
Why
Reduce fragmentation and keep relevant patient context in one place.
What I learned
Therapists need to quickly refresh their memory before sessions.
Decision
Bring previous session summaries and patient context together.
Why
Make relevant history easier to access before and during the session.
What I learned
Therapists spend time creating or finding personalised resources.
Decision
Use AI to help generate personalised resources based on session context.
Why
Reduce repetitive work while keeping the therapist responsible for what is shared.
What I learned
Scheduling and calendar management can be fragmented.
Decision
Explore Google Calendar and Apple Calendar synchronisation.
Why
Make scheduling easier to view and manage from the therapist workspace.

MVP scope

What made it into the product

Therapist

Dashboardfunctional
Patientsfunctional
Patient contextprototyped
Sessionsfunctional
Session summariesprototyped
Agenda / calendardesigned
Resourcesprototyped
AI assistanceprototyped

Patient

Homefunctional
Register a momentprototyped
Progress / historyprototyped
Goalsdesigned
Therapist approved resourcesdesigned
Session related informationdesigned

Transcriptionprototyped
Session summariesprototyped
Insights / patternsconcept
Suggested actionsconcept
Personalised resourcesprototyped
Audio generationconcept

Validation

What still needs to be validated

The MVP has not yet been formally tested with the people I interviewed. The next step is to test the core workflow with therapists and patients before expanding the product further.

Validation questions

Can therapists understand the value of the integrated workflow?
Can they review and edit AI generated content comfortably?
Does the patient experience make it easier to capture useful moments between sessions?
Does the information captured between sessions improve continuity in the next session?
Which parts of the broader workspace are actually necessary?

AI integration

Where AI helps, and where it stops

AI can help with

Transcription
Summarisation
Surfacing patterns
Suggested actions
Personalised resources
Audio generation

Therapist decides

Review
Edit
Approve
Share with patient

AI does not act autonomously

Nothing is sent to the patient without therapist approval.
AI should not replace therapist judgement.
AI outputs should remain reviewable and editable.

AI workflow

AI as a layer inside the loop

Session
AI processing
Summary / insights / suggestions
Therapist review
Edit / approve
Patient
Between session input
Next session

Patient experience

Register a moment

Problem: Patients may experience something important between sessions but forget it by the time the next session arrives.
Design response: Give patients a simple way to capture a moment while it is happening, using voice or text.
Product value: Those moments can become useful context for the next therapy session.
User quote: To be added

Design decisions

Information hierarchy

The therapist needs to understand their work at a glance, while still being able to access deeper patient context when needed.

Home

High level overview of the day and the caseload

Patient

Sessions first
Relevant context
Documentation

Session

Current session
Previous context
AI assistance
Actions / follow up

My role

End to end product design and strategy

End to end product design and strategy, from research and product definition to UX, AI workflows and MVP build.

Responsibilities

User research
Problem definition
Product strategy
MVP definition
Therapist and patient journeys
Information architecture
UX flows
UI design
AI interaction design
Prototyping
MVP build

Outcome

Where Continuum stands today

Continuum is currently at the MVP and prototype stage. The product work helped turn the initial patient problem into a broader product concept connecting therapist workflows with patient continuity.

Current product status: MVP / prototype.
What has been built: A therapist workspace and a connected patient experience, with AI assistance built into the session flow.
What has been validated: The problem space, through 35+ conversations with therapists, coaches and people with therapy experience.
What remains to validate: The core workflow with therapists and patients, and the quality of AI outputs in real use.

Takeaways

What building Continuum taught me

Define the core product problem before expanding the feature set.
Understand both sides of a product relationship.
Ask users about their actual workflow rather than designing from assumptions.
Learn where AI is useful and where human review needs to remain central.
Building with AI makes it easier to explore and prototype, but it does not replace product decisions.

Next steps

Validate the core loop before expanding

Session
AI summary / insights
Therapist review
Personalised support
Patient between sessions
Next session

Potential next validation

Therapist usability testing
Patient usability testing
AI output quality testing
Validation of willingness to use or pay
Identify which wider therapist features are actually needed

Next case study

Unwind

Cutting decision fatigue: an AI-driven companion that recommends movies, music and books based on your current mood.

Read case study