Nick Mann

I design interfaces, build my own apps and mentor designers

Perimeter turns the unused edges of the MacBook trackpad into shortcuts. Take a screenshot or hide a window with one tap, without a key combination.

Product context. I work on a trackpad with a fast cursor speed, so my finger almost never leaves the middle. The edges just stay unused. At the same time, the things I do most often, like taking a screenshot, hiding an app's interface or turning on x-ray mode, were all keyboard shortcuts. Some need four keys at once, and some can't be changed because the app doesn't allow it.

So I moved the shortcuts to the edges of the trackpad. The hand stays where it is.

Year
2026
Type
Self-initiated product · macOS app
Role
Product design, interaction
Status
Released

Solution. Perimeter shows your trackpad in the middle of the screen with points along the edges. Pick a point, press + and add an action: give it a name, choose a trigger (tap, click, force click, swipe or a drawn gesture), then click the shortcut field and press the real combination. Press ⌘⇧4, and this point now takes screenshots. It works with any shortcut, even the ones an app won't let you change.

One point can hold several actions with different triggers. Profiles switch on their own when you change apps, so the same point can do one thing in Figma and another in Safari. An action can also be pinned to all profiles: set it once, and it works the same everywhere.

Fit. Hands are different, so the settings are about fitting the app to yours. Dot size and edge inset change how big each point is and how much of the trackpad it takes. Each side can have its own number of dots, or none. Swipe distance and haptic feedback can be tuned too. All points sit a little away from the edge, so they don't get in the way of macOS gestures.

Restraint. The biggest risk with a utility like this is adding too much. Every new trigger, mode or switch makes it harder to understand. I stopped at points, actions, profiles and the settings that fit them to the hand. Everything else stayed out.

A medical information system for a network of rehab clinics. It took the clinics off paper without changing how the staff works.

Product context. The client is a network of spine and joint rehabilitation clinics, with its own and franchise centers in Russia, Moldova and Turkey. Their specialists have worked on rehabilitation at the Olympic Games and other big sporting events.

The clinics already had their own MIS, and it held something valuable: a patented method doctors use to fill out exam templates. The task was to build a modern system with ERP features on top of it. Keep the method, add what any modern MIS has, and take into account what the client had seen in competing systems over the years.

Four groups work in the system. Reception handles the first contact: requests from the website, phone calls and patient chat, and bookings for the gym, doctor visits, massage and other procedures. Doctors write exam reports for their specialty and attach MRI, CT and other scans. Instructors run physical therapy in the gym using the schedule and the doctor's recommendations. Managers and other non-medical staff use tasks, mail, work schedules, marketing and finance.

Year
2022 – 2024
Type
MIS/ERP for a rehabilitation clinic network
Role
Product design, research, design system
Status
Desktop design delivered and approved
Result
Critical workflows twice as fast

Problem. Four roles with very different work were sharing one system that wasn't made for any of them, and the old architecture couldn't grow anymore. On top of that, by law nothing in a medical record can be fully deleted, and the old system wasn't built with that in mind.

Process. Before starting, we asked administrators and doctors about the current MIS: what bothers them and what they would change. Throughout the project we met with doctors, administrators and clinic heads every week, sometimes twice a week. Every block was discussed with the people who would use it: complaints, exam structure, diagnosis search, templates, scheduling, reporting.

Where the old system already worked, we kept it. The color coding of bookings came from the old MIS, and the client decided to keep it, so nobody had to relearn it.

New screens were checked with people who used the old system every day. We showed a screen, asked how they would do a specific action, and watched whether they found the way. Where there were two solutions, we ran A/B tests with administrators and doctors and kept the one that worked better. If people got lost, the screen was redesigned.

Solution. The interface is built from modules. Parts of it are shown or hidden by role, so reception, doctors and instructors each see their own work. Dashboards and widgets are set up by the head of the clinic: they can be added, removed and changed. Main actions like creating a request or adding a patient are available from wherever you are, not from one start screen.

The palette is neutral and stays within the clinic's brand. Emphasis comes from shape and spacing, not bright color. Every color was checked on TN, VA, IPS and OLED screens and on projectors, because the clinics use all of them.

Reading the exam, not typing it. Instead of searching, the doctor marks zones on a body diagram: nerve root zones, muscle groups, posture, segments of a limb. One visit can use several diagrams. It's easier to find a zone on a figure than to type several items into search one by one.

Nothing is ever fully deleted. By law, data in a medical record can't be erased. When a doctor deletes an entry, they see it as deleted, and the head of the clinic sees it as a crossed-out line. The full history stays in the record.

Templates and diagnoses. Exams have a lot of repeated text, so doctors can build a template for any diagnosis: complaints, fields, recommendations, restrictions and body diagrams. During the visit the doctor can edit the template or type by hand. Diagnosis search works by name and ICD code, and the most used diagnoses come first. This replaced the thickest stacks of paper in the clinic.

Result. The client wanted to see mockups of the main parts before starting development, and after seeing them they decided to go ahead. Critical workflows now take half the time they used to, and staff kept their usual way of working. The design system was simple enough that the front-end developer put together the whole component library in a week. Now the developers are building one web app for all platforms, and next come the clinic head's block, reporting and analytics, and integration with accounting.

A mod manager for GTA V, redesigned around what its engine can actually handle.

Product context. FiveM is a multiplayer platform for GTA V. FiveMods, made for OpenCore, is the app where players find, install and manage mods. The old app had hit its limits: the structure couldn't fit the new features and offers, and the look the client wanted was too heavy for the app's render engine. I worked on the redesign together with developers, front-end engineers, SEO specialists, marketers and product managers.

Year
2021 – 2025
Type
Consumer software · Mod manager for GTA V FiveM
Role
Product design, design system, brand identity
Status
Released
Result
2,000,000+ app downloads

Problem. The brief asked for a modern interface with blur and layered panels. But real-time blur was too heavy for the app's render engine and cost frame rate the product couldn't spare. The analytics showed one more thing: players almost never opened the app full screen, so a fixed desktop layout would have been designed for a mode nobody uses.

Process. First I tried the obvious way: live blur across the whole interface. It cost too much frame rate. Then I looked at what blur looks like, not how it's calculated. A blurred background seen through a semi-transparent panel looks the same as a blurred panel. Once transparency did the job, the heavy version was dropped.

The layout went through several rounds. Each one was checked against user feedback and how the previous version actually performed, not only against the brief. That's where the floating window came from: players rarely went full screen, so the layout was rebuilt around the window size they really use.

Solution. The interface is built from modular panels. Each panel does one job and can be replaced or moved as the product grows. Blur is applied once, to the background, and everything on top uses semi-transparent color, so it looks blurred for free. The interface got its depth, and the engine kept its frame rate.

Every screen was designed as a floating window first, so the layout works at any size and resolution.

Beyond the app. The work went beyond the app itself: a new brand identity from scratch, a companion desktop app and the company website.

Pointer is a mouse built around one idea: the actions you repeat every day should feel natural.

Product context. We use dozens of shortcuts every day. Some can't be changed, some we don't want to change, and some make you take your hand off the mouse.

Pointer tries another way. Gestures, pressure, mouse movement and touch can all become inputs. Draw a circle to hide the interface. Use movement gestures for complex actions. Press under your thumb for FN functions.

Year
2023
Type
Hardware and software concept · Mouse
Role
Industrial design, product design, interaction
Status
Concept

Hardware. The shape is inspired by 80s car design: sharp surfaces on a familiar shape. The body is split into two parts on purpose. When you press harder, the shell gives a little, and a vibration motor recreates the feeling of a mechanical switch.

Software. The app stays out of the way. Profiles switch on their own depending on the active app. Settings are stored in the mouse itself, so the web app can set it up without running in the background. Global gestures can be shared between profiles, so you don't set them up twice.

Even small things can be adjusted. Tap zones fit the size of your finger, a gesture fires when your finger leaves the zone, and scrolling and haptic feedback can be tuned to your liking.