PRODUCT MANAGER · 2027 GRADUATE / Cloud-Native Live

KOREY WANG TIANRU · PRODUCT MANAGER

Three tools built from recurring friction: TEM-4 Workspace, Fitness OS, and the DM Runbook.

When studying, checking answer keys was scattered across folders. When coaching gym members, notes lived in separate apps. On stage, thick binders were useless under live stage pressure. I start with tangible friction on the ground, then figure out the layout, the documentation, and where human verification must guard against automated guesses. AI writes the code; I own the problem, the decisions, and the real-world testing.

VERIFIED PRODUCTS
LIVE PROTOTYPES
tem4-workbench.local · Revision Loop
TEM-4 Study Workspace Demo
9 Yrs Past Papers · 627 Question Pool · Mistake Diagnostics View Case →
Fitness OS iPhone Prototype
Interactive Prototype Live
HealthKit Integration + Local Storage + 7d Cycles Watch / Open product ↗
dm-console · Performer Rehearsal
DM Rehearsal Console
132p Prep Guide · 34p Live Runbook · 9-Act Stage Cues View Case →
Start on the floor Keep data local Know who uses it Iterate after use Honest on progress

SELECTED WORK

Three projects, all born from friction I ran into myself.

Every case includes real interfaces, interactive prototypes, and verifiable deliverables. Source code remains local and private; public pages feature sanitized assets and live interactive demos.

TEM-4 Study Workspace · Feedback Loop
TEM-4 Study Workspace Interface Sanitized interface from the actual local study desk
01 WEB PRODUCT · EDTECH

TEM-4 Study Workspace

After answering a drill, what should you practice next?

During exam preparation, my practice questions, answer keys, and error notes were scattered across notebooks and web tabs. I consolidated them into a unified desk: answer questions, submit to see answers and grammatical breakdowns, and automatically route errors to a review deck. Later, I replaced hardcoded dates with customizable target scores, closing the loop from practice to review. Official TEM-4 scores rose from 53 to 69 during use.

180
9 Yrs Past Grammar
627
Expanded Practice Pool
13
Dictation Drills
Fitness OS iPhone Today View Native iOS architecture + Interactive Web Demo
02 iOS PRODUCT · WEB PROTOTYPE · HEALTH

Fitness OS

After logging workouts and meals, how do you easily review the trend?

Both in my own training and when coaching gym members, I needed to review workouts, nutrition, and subjective fatigue in one place. I built an app around that workflow: enter raw notes, verify the parsed date and content before saving, and revisit history on a clean timeline. Unknown values are never stored as zero. The native app pairs with Apple Health to suggest minimal daily actions, supported by an interactive web demo for zero-friction evaluation.

28d
Personal Baseline Window
7d
Single Action Cycle
Local First
Privacy & Zero Zero-Fills
DM Rehearsal Console · Synthetic Script
DM Rehearsal System Overview Synthetic script demonstration interface
03 AI WORKFLOW · LIVE OPS

DM Rehearsal & Role Handbook

Before stepping on stage, how do you see at a glance who to cue, what to do, and which line to speak?

While working as an immersive role-play game host (DM), I compiled a 132-page rehearsal manual for deep backstage prep and a 34-page live runbook for on-stage cue reading. Information is structured by act into entrance signals, physical actions, and exact lines. The venue owner shared the guides with other hosts as standardized reference. Complex narrative material compressed into an actionable live index.

132
Rehearsal Manual
34
Live Stage Runbook
9
On-Stage Act Cues

HOW I WORK

How I turn raw friction into practical tools

01

Start on the floor

When coaching, studying, or performing, note the exact step where things grind to a halt before deciding on features.

02

Know who uses it

Focus the screen on immediate needs. Grounded in practical validation: where attention comes from, who finds real value, and how the first loop holds up.

03

Let AI write the drafts

Write clear scopes, protect essential content, and specify acceptance checks before handing execution to AI coding tools.

04

Test it with my own hands

A page opening without errors is just the beginning. I walk through the flow step-by-step as an actual user, checking edge cases, empty states, and save behavior.

05

Be honest on progress

Running locally, public demo, tested on device, and validated by users are distinct stages of maturity.

LOOKING FOR

Product Manager focused on real user problems

Focusing on AI applications, human-in-the-loop workflows, user empathy, and bilingual project coordination. To explore the design process in depth, check out the case studies or download the 4-page PDF.