Jay ParkProduct design · Systems & craft
Selected work04 / 08

Peptide Library

01

From fragmented research to a shipped iOS app.

Fragmented research became a source-backed library, then a tracker you can keep using.

My role
Product Designer
Scope
Product direction · UX & visual design · Native implementation · QA
Delivery
Live on the App Store, version 1.2

Interactive prototype

A web sample of the app.

Peptide Library app icon

Explore

Phone recreation

02

Information was everywhere. Reliable answers were harder to find.

Studies, threads, and conversations each held a piece. None of them held the whole question.

Fragmented research

A study is a primary source. A forum thread and a gym conversation are public discussion. Questions sit with them: what it does, the risks, the evidence, what has been studied, what it could help with, and how reliable a claim is.

  • Primary sourceA study, with an abstract and a result on its own.
  • Public discussionA forum thread where people compared what they tried.
  • Public discussionA name passed from person to person at the gym.

Public discussions showed where the questions were. Primary sources are what the compound pages use as evidence.

  • What does this actually do?
  • What are the risks?
  • Is there evidence?
  • What has been studied?
  • What could it help with?
  • How reliable is this claim?
How I did this

Public conversations were the discovery signal: feeds, threads, gym circles, and people around me. Primary sources are the evidence on the compound pages. I reconstructed the session for this case study. Themes were synthesized with AI help. No survey export survives.

03

Different questions revealed four core information needs.

This is for someone comparing a name from a study, a thread, or the gym. I listened in public: feeds, Reddit, gym circles, and people around me. The lines below are paraphrased. I reconstructed them for this page. There is no survey and no count.

Questions grouped by need

Paraphrased questions grouped into four needs: What is it, Evidence, Risk, and Outcome.

Retrospective reconstruction

  • What is it

    What even is this compound?

    Is this approved, or still investigational?

  • Evidence

    Is there human evidence, or only animal work?

    Where did this claim come from?

  • Risk

    What are the side effects?

  • Outcome

    Is this something people use for weight?

Paraphrased questions, grouped into four needs.

04

Four user needs shaped the compound page.

Each need maps to a section of a real compound page.

Semaglutide compound page

Mapped from the shipped page

The four needs connect to regions of the shipped Semaglutide page: what it is, what it could help with, the risks, and the evidence.

Semaglutide

FDA approved

Illustration of a frosted glass vial

A medicine that acts on the GLP-1 pathway.

GLP-1 receptor agonist

Commonly explored for

  • Illustration of a balance scale and an appleWeight lossFDA Approved
  • Illustration of an appleBlood sugarFDA Approved

How it works

  • Appetite controlActs on pathways involved in appetite regulation and can reduce food intake.
  • Metabolic effectsAlso studied for effects on blood glucose and other metabolic measures.
  • GLP-1 pathwaySemaglutide acts on the GLP-1 pathway.

Good to know

Digestive side effects are common

Serious warnings apply

Safety

Potential side effects

  • Nausea
  • Vomiting
  • Diarrhea
  • Constipation
  • Abdominal pain

Warnings

  • Not for everyone. Contraindications and boxed warnings are in official labeling.
  • Talk to a licensed clinician before starting, stopping, or changing any medicine.

Evidence

Established clinical evidence (labeled uses)

Large human trials and regulatory reviews support labeled uses. This is not a complete list of indications and is not a recommendation.

Sources

Last reviewed: 2026-08-29

05

The first build proved the structure. The next gave it clarity.

This is Semaglutide in both builds.

Semaglutide, first build and later polish

Screenshots

Semaglutide in the first build beside Semaglutide after the visual polish. On a small screen, the two screens switch a single frame.

Aug 30. Semaglutide.
  • IdentityThe later screen puts the name on a navy hero, with a frosted vial. The first screen is a paper page, teal type, and no vial.
  • HierarchyThe name becomes the large line. The sections under it stay smaller, so the page scans.
  • StatusFDA approved stays with the name, in a pill on that hero.
06

A consistent visual language for complex information.

Category marks, compound vials, and the app icon each follow their own rules.

Visual system

Art-direction boards

Two art-direction boards, then three rules: category marks, compound objects, and the app icon. Five category marks include Recovery and Clinical Studies. Two TestFlight frames show the marks in the product. They are not version 1.2.

How I did this

The boards are the production art direction for the category marks. Compound vials and the app icon follow different rules, because they do different jobs. The illustrations arrived with the visual polish. The phones are that TestFlight build, not version 1.2.

07

Designing, building, and validating on the actual device.

I checked changes on the running app on device as I went.

Checked on the running product

Checked on device

Dark Mode on Check-in: facial details were hard to see, the relevant details moved to adaptive treatment and colors, and the corrected state was checked in the running product. Then a loop of design, build, run, evaluate, and refine.

How I did this

I worked in the native build and kept the product calls. Eighteen commits through Sep 16. Fourteen were co-authored with Cursor. The Dark Mode fix is e6c59ee. Build 1 went to TestFlight. On-device testing exposed a Dark Mode contrast issue in Check-in. The Build 1 failure was not captured. Build 2 is the Sep 4 upload.

08

A reference library answered questions. Tracking gave people a reason to return.

Why tracking came next

Paraphrased themes from informal feedback. The reference loop is question, research, learn, leave. The recurring loop in 1.2 is research, plan, track, reflect, return. Under that: 1.0 was the library plus a lightweight journal, 1.1 improved check-ins, and 1.2 added Today, Calendar, Cycle, and Stack.

09

Learning and personal tracking, connected without crossing the line.

Compound pages teach, the tracker keeps what the person typed, and neither gives dosing advice.

Library and tracker

Explore and compound pages sit on a reading surface. Today, Stack, and Check-in sit on a personal surface, split by a hard line. Research can inform reflection. A compound page never becomes a treatment plan. A short note says what to measure next: return frequency, repeat check-ins, and source engagement.

Explore

Semaglutide

GLP-1 receptor agonist

FDA approved

  • Commonly explored for
  • Evidence
  • Sources

Compound pages

Today

Weekend cycle

Open

Semaglutide

0.25 mg on the schedule Not a recommendation.

  • Check-in
  • Stack

What the person typed

Research can inform reflection. A compound page never becomes a treatment plan.

What I'd measure next: how often people come back, whether check-ins repeat, and whether they open a source.

10

From the first build to a live App Store release.

These dates are the build record.

Version 1.2 release

The build record, then Peptide Library on the App Store.

  1. Aug 30First build
  2. Sep 3Visual polish
  3. Sep 4Build 2 to App Store Connect
  4. Sep 8Version 1.0
  5. Sep 9Version 1.1, improved check-ins
  6. Sep 17Version 1.2, 9:32 AM PT
Peptide Library on the App Store

The App Store icon.

Next / MonkeyTilt

I designed and shipped Crews, achievements, and challenges.