ANCHOR

Preventing relapse through real-time, AI-powered support

A mobile app designed to prevent relapse by offering real-time, personalized coping tools and support when users need it most.

A mobile app designed to prevent relapse by offering real-time, personalized coping tools and support when users need it most.

A mobile app designed to prevent relapse by offering real-time, personalized coping tools and support when users need it most.

YEAR

2025

DURATION

4 Weeks

Role

Sole UX/UI Designer

YEAR

2025

DURATION

4 Weeks

Role

Sole UX/UI Designer

BACKGROUND

Existing Recovery Apps Fail When They're Needed Most

Imagine this: It’s 11 PM. Jamie, a person in recovery, just finished a stressful workday. Anxiety is high, and the urge to drink creeps in. They open a sobriety app—only to see a streak counter, not the immediate support they need.

"85% of relapses occur during high-risk trigger moments when real-time support isn’t available."

Journal of Substance Abuse Treatment, 2022

Journal of Substance Abuse Treatment, 2022

As someone with lived experience in recovery, I know firsthand how a split-second decision can be a matter of life or death. That’s why I created ANCHOR—a recovery tool designed to provide immediate, AI-powered assistance when it matters most.

THE PROBLEM

People navigating recovery need accessible and immediate support during high-stress situations to prevent relapse

THE SOLUTION

ANCHOR: Bridging the Gap with Real-Time Support & AI Assistance

RESEARCH

Leading Discovery with Empathy

To validate my hypothesis, I explored existing recovery apps and interviewed individuals in active recovery.

Given the deeply personal and sensitive nature of substance use recovery, I approached research with empathy and care. Establishing trust was key to gathering honest, meaningful insights that shaped the direction of ANCHOR.

To validate my hypothesis, I explored existing recovery apps and interviewed individuals in active recovery.

Given the deeply personal and sensitive nature of substance use recovery, I approached research with empathy and care. Establishing trust was key to gathering honest, meaningful insights that shaped the direction of ANCHOR.

Competitive Analysis: Strengths & Gaps in Existing Solutions

I reviewed four major recovery apps and uncovered key insights:

Strengths
  • User-Friendly & Clean Design

  • Comprehensive Sober Tracker

  • Supportive Community & Daily Motivation

  • Location-Based Meeting Finder

  • Evidence-Based Approach with Resource Database

Weaknesses
  • Limited Personalization

  • No Real-Time Support

  • Overwhelming Content for New Users

  • Narrow Target Audience

  • Requires In-App Purchases for Personalization

User Interviews: Uncovering Core Needs

I conducted four in-person & virtual interviews with individuals at various recovery stages, from newly sober to long-term maintenance. Using an affinity map, I identified the following insights:

User Interviews: Uncovering Core Needs

I conducted four in-person & virtual interviews with individuals at various recovery stages, from newly sober to long-term maintenance. Using an affinity map, I identified the following insights

DEFINE

Rooting Design Decisions in Real User Needs

To ensure my design was grounded in the needs and challenges of individuals in recovery, I created a user persona based on key insights from my research.

Understanding Our User: Meet Jamie!

Jamie needs a tool that offers real-time, judgement-free support and personalized guidance during high-stress moments.

HOW MIGHT WE…

Provide real-time, judgment-free support to individuals in recovery during moments of high stress to prevent relapse?

IDEATION

Turning Pain Points into Features

With a clearly defined user, I explored solutions that align with their needs and challenges to ensure support is always available.

Why AI?

AI provides immediate, personalized support by addressing key pain points:

  • Lack of real-time guidance → AI offers adaptive, real-time recommendations based on user history and triggers.

  • Limited availability of human support → AI enables 24/7 crisis assistance without fear of judgment.

  • Generic recovery resources → AI delivers personalized suggestions tailored to each user’s journey.

Information Architecture

To align my solution with user needs, I first created a site map to ensure seamless navigation. I then designed user flows that prioritized quick, intuitive access to personalized support, based on key research insights:

  • AI Recovery Chat + Meeting Join: Users interact with the AI and receive meeting recommendations.

  • Crisis Support + Post-Call Check-in: Users can quickly access emergency support and log their emotional state afterward.

DESIGN

Designing for Support & Accessibility

After defining core user flows, I sketched key screens to visualize interactions, prioritizing a clean, accessible design to reduce cognitive load. Given the app’s sensitive nature, the goal was to minimize friction and ensure quick access to support.

From there, I transitioned to Figma, refining layout and hierarchy with mid-fidelity wireframes before adding visual elements.

Visual Elements that Drive the Solution

Every element of ANCHOR’s UI was crafted to evoke stability, support, and hope, ensuring a seamless and reassuring experience.

  • Logo & Colors: A minimalist anchor symbolizes grounding support, paired with soothing blues and neutrals for reassurance.

  • Typography: Inter was chosen for its clarity and warmth, maintaining readability without feeling clinical.

  • UI Components: Large CTA buttons, intuitive icons, and structured layouts reduce friction, making navigation effortless.

This cohesive design approach ensures ANCHOR remains accessible, inviting, and easy to use—even in moments of uncertainty.

TESTING & ITERATIONS

Refining for Maximum Impact

I conducted a remote, moderated usability test with five participants to evaluate key flows within ANCHOR. They were asked to complete the following tasks:

  • Interacting with AI Chat and joining a meeting.

  • Accessing crisis support and completing the post-call mood check-in.

While the app performed well overall, feedback highlighted opportunities to enhance visibility, accessibility, and tone. Iterations were made to refine the user experience accordingly.

RESULTS

With an average satisfaction rating of 9/10, users found the app to be especially valuable for individuals in early recovery

The app received a strong 9/10 average satisfaction rating, with users highlighting its value for individuals in early recovery. 80% of testers said they would use the app regularly, reinforcing its potential to provide ongoing support during critical moments.

"This would have helped me so much in early recovery." – User Feedback

"This would have helped me so much in early recovery." – User Feedback

AI Recovery Chat & Meeting Suggestions

Crisis Support & Post-Call Check-in

FINAL THOUGHTS

What I Learned: This project highlighted the importance of empathy in design. Understanding the emotional and psychological needs of users in recovery influenced every phase, from research to design. Balancing these needs during vulnerable moments was a challenge, but it deepened my ability to create supportive, user-centered solutions.

Next Steps: In the future, I’d focus on expanding the community feature to foster deeper connections between users. This would create a stronger sense of belonging and support, which is critical for people in recovery. Additionally, refining AI personalization based on user history would further enhance the app’s ability to meet individual needs.

What I’m Most Proud Of: I’m incredibly proud to have designed a solution that has the potential to save lives. ANCHOR is more than just a tool for tracking sobriety—it provides real-time, stigma-free support for individuals navigating recovery. Seeing how this solution can positively impact users in their most vulnerable moments is something I truly value!

LIKE WHAT YOU SEE?

Say Hello!

Linkedin

Matt Fredette 2025

2025 Matt Fredette

LIKE WHAT YOU SEE?

Say Hello!

Matt Fredette 2025

2025 Matt Fredette

LIKE WHAT YOU SEE?

LIKE WHAT YOU SEE?

Say Hello!

Say Hello!

Matt Fredette 2025

2025 Matt Fredette