Mohamed Mufeeth Product designer
Hire me

Open to product design roles and freelance work

MohamedMufeeth

Product designer

I spent six years running delivery operations. Now I design products and ship them.

Apps and watch faces live on Google Play, a watch face platform, a team of AI agents that runs my releases and about 65% of dispatch automated at Kitopi.

Message on WhatsApp Resume

Based in Tamil Nadu, India, open to relocate

A scale model of where it started

Six years on the operations floor.

At Kitopi I ran last mile delivery for a cloud kitchen. Several brands cooked under one roof, and well over a thousand orders a day went out of the door. I learned how products behave when real people depend on them.

I kept seeing what broke.

Riders waited, ready orders sat on the pass, and nobody could see why. So I started designing the fix. First in plain words on the kitchen floor, then as rules written from the office, and now as an independent product designer.

Kitopi, on the floor as fleet supervisorShipped

OCR Rider Check In

Nobody could see who had arrived.

Riders from several delivery apps came in all day. One dispatcher typed every arrival into Excel by the last few digits of the order number. It was too much for one person, and orders went to the wrong rider.

So I said the fix out loud.

“What if there's a scanner here, and when the rider comes, he scans it, and it shows up on the TV, so the kitchen knows, the dispatcher knows, and the order goes out smoothly.”

Said to our product owner, in words close to these. The team built it: the rider scans the order number, and it shows on a TV with its status for the kitchen and the dispatchers.

No arrival scan, no dispatch.

Only an order that is ready and has its arrival scan can be handed over. Hand one to a rider who has not scanned, and the system stops it. Wrong dispatches fell, every arrival is logged and fewer staff run the counter. I spotted the problem and pitched the fix; the product owner and the design and engineering team built it. Kitopi later turned it into self pickup from a shelf.

Illustration of the flow, not the production system

Kitopi, from the officeShipped

Dispatch Automation

Every order waited for a person.

I had moved from the kitchen floor to the office, in customer experience and last mile operations. There, coordinators assigned almost every order by hand in Onfleet. At peak, orders waited while riders sat idle, and the overflow went to costly third party fleets.

Two minutes, then the system decides.

If nobody assigns an order within two minutes, the system checks who is free, who is nearest to the pickup and who covers the zone, then assigns that rider.

Safe batching, and a next step for every failure.

Two orders share a rider only when pickups are about two minutes apart, drops within three, and the delivery time still holds. If all riders are busy, the order goes to a third party fleet. A rider who does not start gets a ticket to an agent.

about65%

of dispatch automated, from fully manual.

The operations team got over 40% smaller while delivery times and customer satisfaction held. I owned the rules, worked them out with the product owner, and they were built into Onfleet.

Illustration of the rules, not the production tool

Self initiated while I studied UX and UIPrototype

Group Ordering

One person ordered for eight.

Team lunches ran through one organizer. Collect every choice across notes, chats and the ordering app, retype each item, check the total by hand and hope. Nobody else could see if their item made it in.

I started with the people who order together.

I interviewed a university student, staff support, a manager, employees and social organizers, then mapped what I heard into an empathy map, a journey map and How Might We questions. Two people led the design. Laila wants to enjoy the outing, not run it. Victor is allergic to dairy and worried his choices get missed.

One link instead of one organizer.

The organizer sets the kitchen, the deadline and a budget per person once, then shares one link. Everyone adds their own items. No account needed: you pick your name from the organizer's list.

The budget is the interface.

Each person lands on whose order it is, the deadline and their budget, with a live meter of what is left. The organizer sees who joined, who is ordering and who is done, then places the order in one tap. Testing turned a generic start button into a group status banner.

5to1

manual steps for the organizer became one setup step.

In testing, every organizer finished setup and shared the invite link without help. I found the gap myself and took it from interviews to a tested hi fi prototype in four weeks. Only after I finished did I learn that Kitopi's product team was building the same feature.

Illustration of the flow, not the prototype screens

Then I started building my own.

I started Design Blues in 2023 next to my operations job. In July 2025 I left Kitopi, came home to India and went independent. Now I look for problems worth solving, then spec, design, build and ship the answers, using AI coding tools for the engineering.

Founder, product and UX designOnboarding creators

WatchFace Explorer and CreatorOS

Great watch faces get lost.

I design watch faces myself, first for Samsung's Galaxy Store and then for Wear OS. On the Play Store, every face is one listing among many.

A listing is not a creator.

No name to follow, no audience to keep, no way to tell past fans about a new face. Ranking needs installs and installs need ranking. Every release starts from zero.

So creators got a home of their own.

WatchFace Explorer gives each creator a profile, followers and a one tap Open on watch. Google Play still handles the install.

Drops that creators control.

Coupon drops are built in. Each claim is tied to one person and one face for good, a rule a real production bug taught me. Notifications have a budget and a cooldown, so fans are never spammed.

Three apps, one platform.

Explorer for fans, CreatorOS for creators and Core for moderation, live since August 2026 at watchfaceexplorer.com and on Google Play. I am onboarding the first creators now.

Visit the site

Real watch faces and real screens; laptop and tablet screens are illustrations

Designed, specced and built by meLive on Google Play

NiNav: Subscription Tracker

Subscriptions renew on their own.

Streaming, cloud storage, the gym, a free trial that turns into a charge. Each one renews on its own date, whether you still use it or not.

And they are not the only dates.

Bills, insurance, warranties, EMIs and passports come with dates too. They end up in a drawer, an inbox or some app, and nobody remembers where.

And the dates slip by.

It happened to me. Bank cards, a passport and paper documents expired before I noticed, and I could not remember where I had saved half of my subscription dates.

Track every subscription, and every date that renews.

Type or say “Car insurance renews 12 March”, or scan the bill. NiNav works out what it is and when, and keeps it with everything else you track.

It tells you before it renews.

Overdue, due soon or upcoming, most urgent first, with a calendar of what is coming up to pay. Specced in seven documents, then designed and built by me. Live on Google Play since October 2026.

Get it on Google Play

Real app screens on the phone

Designed by me, built with AI coding toolsIn daily use

Design Blues HQ

One person, too many releases.

Design Blues is me alone: watch faces, their phone apps, NiNav and WatchFace Explorer. Every update needs a design, a build, testing, store images, a listing and a release. I was the queue.

So I built a studio of AI agents.

In HQ each product has its own office: a product lead, a designer, an engineer, QA, store listing, media and support. Each one is an AI agent with a written brief, a model chosen for the job and a spending limit. I talk to Kip, my main agent, or to an office's lead.

Heavy work on my PC, the rest in the cloud.

Builds, emulator tests, pictures and films run on my PC. Chats, listing writing, translations, review reply drafts and weekly reports run in the cloud, also while the PC is off. Work that needs the PC waits in a queue until it is back.

Nothing goes out without my OK.

Every release goes to Internal testing, then a security check, then Production on my word. Review replies stay drafts until I send them. Build cannot publish, Test can reach only its two emulators, and the release is a script with no AI in it that checks my approval again.

It runs my studio every day.

Updates to my watch faces and apps go through HQ, from my phone, my PC or the web. I designed the offices, the rules and the app, and built it with AI coding tools.

Illustration of how HQ is organised; the phone shows HQ screens from October 2026

Bespoke digital horologyOn Google Play

Design Blues watch faces

Many watch faces are just pictures.

A flat image with the time on top. I wanted mine to look and behave like real instruments. This is Flight Horizon, a cockpit attitude indicator for the wrist.

Built in layers, like the instrument.

A horizon at the bottom, the pitch ladder and bank scale over it, then a panel with heart rate, time and steps on tapes with rolling digits. Each layer moves on its own.

It moves with your wrist.

Tilt your wrist and the horizon banks and pitches. Five colour styles, three horizon styles and four always on styles let people make it theirs. Console Flight is built the same way.

Every face is a package.

Each face comes with its own phone app: widgets, a live wallpaper and a still one, and a bedside clock while charging, all in the face's own colours and type, never stock Android.

A face cannot read every sensor. So I built Almanac.

Watch faces cannot run code, so they cannot do their own maths. DB Almanac is my compass, weather and sky app, and it feeds my faces live data such as true north, sun and moon position and golden hour.

A real face on a 3D watch, first rendered, then live; the phone shows the real app

Operations taught me how products behave when real people depend on them. Design is how I fix what I see.

Real constraints first

I start from a real constraint, not a hypothetical user.

Design for after launch

Guardrails, edge cases and failure modes, not only the happy path.

Real over impressive

I show what is real, and I say what I did not build.

  1. January 2019 to July 2025Kitopi. Fleet supervisor on the kitchen floor, then customer experience and last mile operations from the office: rider management, process improvement and customer handling.
  2. 2023Started Design Blues next to my job: watch faces for Galaxy Store, then Wear OS.
  3. January 2024Completed the Google UX Design Certificate, with K-Fleet as its practice project. From then, six hours a week with Kitopi's product team, next to my operations job.
  4. July 2025Left Kitopi, came home to India and went independent.
  5. August 2026WatchFace Explorer live.
  6. October 2026NiNav live on Google Play. Started Design Blues HQ, my studio of AI agents, now in daily use.

Your product is next.

Open to product design roles and freelance projects. Based in Tamil Nadu, India, open to relocate. WhatsApp is the fastest way to reach me.