I design products aroundhow people actually think

Background in cognitive science and psychology. I run the research that shapes the product strategy, then design the system that ships it across consumer and B2B platforms.

Selected work

Password-protected case study.
- Web + Mobile
- B2B SaaS
- UX Research
AI Brokerless Freight Platform
The freight platform built for carriers, not dispatchers
I designed one record carriers build from their own actions, which shippers and admins read live.
- In-dev · venture-backed
- 25+ carriers & shippers in pilot
Web + Mobile
B2B SaaS
UX Research
The freight platform built for carriers, not dispatchers
Shipper · Shipment Visibility
Carrier · Nav record
2025
7 months (phase 1)Phase 2 ongoing
Role
Product Designer
Team
2 product designers + 2 PM + 2 eng + 1 director
Scope
Onboarding · carrier actions · shipment visibility · admin · payouts
TL;DR
91.5% of US carriers run 10 trucks or fewer: no back office, but the software assumes one. From 5 carrier interviews and 6 competitor analyses, I built the record from the carrier's own actions: confirm the shipment, file the proof and cost, generate the invoice. Shippers and admin rely on that same record live.
25+pilot
Carriers and shippers
AI Brokerless Freight Platform case study is password-protected
Don't have the password? Copy my email and ask for access.
Password
- Web
  B2B + B2C
  UX Research
  NKOMOR
  Pharmacy workflows that secured 3 partnerships in 60 days
  I shipped the pharmacy tools for launch, products, and orders, and designed the storefront and checkout.
  Partner-approved · shipped
  Web
  B2B + B2C
  UX Research
  Pharmacy workflows that secured 3 partnerships in 60 days
  Operator · Orders MVP
  Consumer · Pharmacies-first
  2024
  May–Aug · 4 months
  Role
  UX/UI Designer
  Team
  2 designers + founder
  Deliverable
  Pitch prototypes → my scope live
  TL;DR
  NKOMOR is a 2-sided pharmacy platform for Africa: customers buy online, operators run the store. I interviewed operators and mapped 6 workflows from what they already ran: spreadsheets, fax, inventory, orders. Then cut to 3 for the pitch. At 38% internet penetration in Africa, recent adopters needed familiar screens, not novel ones. I shipped launch, products, and orders; designed 2 storefront directions and checkout.
  3partnerships in60days
  Pharmacies
  Solution · Operator
  Pharmacy tools that looked like work partners already knew
  Goal gradient effect
  People push harder toward a goal when they can see how close the finish is.
  Launch your store in about 20 minutes
  5 products to go live, with progress visible the whole way. Even an operator who had never used software like this could see the end from the start.
  Recognition over recall
  Matching an existing mental model costs less than teaching a new one.
  Tables: Operators still lived in spreadsheets and fax
  Upload, edit, delete in a CSV table, not a modal maze. Inventory had to feel like the workflows they already ran on.
  Information Foraging Theory
  Users scan pages looking for "scent cues" (like a bright "Mark Shipped" button) that yield the highest reward for the lowest amount of physical or mental effort.
  Orders, 1 click from status to done
  Status list with 1-click actions: mark shipped, sync inventory. Pending and cancelled showed in the MVP list, enough to triage without a full exception-design pass.
  See full case study
- Web + Mobile
  B2C
  UX Strategy
  Job Search Marketplace
  Onboarding that earns completion instead of forcing it
  Designed onboarding from scratch: 1 path with Save & Exit after cutting 3 modes that only changed how the ask felt.
  ↓ 35% drop-off
  ↓ 40% time to apply
  ↑ 32% questions done
  Web + Mobile
  B2C
  UX Strategy
  Onboarding that earns completion instead of forcing it
  Desktop · Save & Exit
  Mobile · Video answer
  2025
  ~3 months
  Role
  Sole UX/UI Designer
  Team
  CEO + PM + 1 eng
  Scope
  Onboarding · dashboard · unlock
  TL;DR
  Job search marketplace onboarding, designed from scratch. iCIMS (2025): 60% started an application and never finished. Sign-up forced full onboarding before users saw the value. I cut 3 urgency modes into 1 path with Save & Exit; the modes only changed how the ask felt, not what we needed to collect.
  ↓ 35%
  Drop-off
  ↓ 40%
  Time to first apply
  ↑ 32%
  Required questions done
  Solution
  Start informed, leave mid-flow, finish for the payoff
  Reason-giving
  People finish more when they understand why you're asking, not just what to fill in.
  Start informed: video walks them through the why
  A video before the questionnaire explains why we're asking. Value before the profile wall.
  Zeigarnik effect
  Unfinished tasks stick in memory. Visible progress gives people a reason to close the loop.
  Leave anytime: progress follows you
  Progress shows on the dashboard, so people have a reason to come back. Leaving mid-flow was part of the design: built for return visits, not first-session completion only.
  Premack principle
  A reward people want (matches, faster apply) makes finishing the profile more likely.
  Finish unlocks matches and faster apply
  Complete the path and you get personalized matches plus faster apply. That's the reason to come back.
  See full case study
- Mobile App
- B2C
- 0→1
LOOKR
Photo inspiration you can use while you shoot, 0 to 1 in 4 weeks
Led research, design direction, and the visual system for a camera app shipped to real users in one sprint.
- Led a team of 2
- 72% adoption intent
- Investor-approved
Mobile App
B2C
0→1
Photo inspiration you can use while you shoot, 0 to 1 in 4 weeks
Camera · tap to generate inspo
Library · export off camera roll
2023
4 weeks
Role
Lead Product Design Intern
Team
2 interns + CEO
Scope
De facto PM
TL;DR
The founder had 4 weeks and a blank brief. I interviewed creators, merged 2 research paths into 1 MVP persona, and cut the sprint to what we could ship. Creators saved inspiration on Pinterest and Instagram, then failed to match it on location. I put Gen-AI references in the camera at the shoot; when overlay-on-viewfinder failed in testing, I moved inspiration below the frame.
72%
Adoption Intent
Solution
Real-time inspiration tailored to where you're shooting
Distributed cognition
Generated from this spot, not recalled from another board.
Shoot, pick what fits, pin it in the frame
Community inspiration. Curated feed to save references.
Compare photos. Takes against pinned inspiration.
Gallery + export. In-app library. Export off the camera roll.
See full case study

About Helen Li

I'm Helen. I got into design helping my immigrant parents with technology and realizing the same interface lands completely differently depending on who's using it. Since then, I've designed for how real people actually think, not how we assume they do.

Education: UC Santa CruzB.S. Cognitive ScienceB.A. Psychology

Labs

Paste a recipe link → Reformatted to be able to cook hands-free
Loading, 0 percent
- WIP
- Consumer
- Cursor
Paste a recipe link → Reformatted to be able to cook hands-free
Savr — paste a recipe link into step-by-step hands-free cook mode
- Healthtech
- Figma
Your care info, finally in one spot
MediLog — appointments and records in one place

Playground

I'm curious and always building. AI experiments in 3D, textures, motion, and UI.

I design products aroundhow people actually think

Selected work

The freight platform built for carriers, not dispatchers

Pharmacy workflows that secured 3 partnerships in 60 days

Onboarding that earns completion instead of forcing it

Photo inspiration you can use while you shoot, 0 to 1 in 4 weeks

About Helen LiAbout

Labs

Paste a recipe link → Reformatted to be able to cook hands-free

Your care info, finally in one spot

Playground

About Helen Li