I don't separate design from making. I sketch ideas directly through prototypes, test them quickly, and iterate until the product feels sharp. The result is software that helps people move with less friction and more confidence.
Read more about my practice
3 projects – spanning healthcare systems, internal tools, and consumer products.
1
BeerMeHelping craft beer bars turn menus into interactive experiences.
Read More
2
ShiftrAppTurning a single bar's scheduling chaos into a multi-tenant operations platform for hospitality teams.
Read More
3
Meuhedet ReimaginedA concept redesign of Israel's second-largest health fund member experience, focused on clarity, trust, and faster task completion.
Read More
Helping craft beer bars turn menus into interactive experiences.
01 – Background & Challenge
While working at a renowned craft beer bar, staff managed over 130 rotating beers. Printed menus quickly became outdated as the selection changed, creating friction for guests and forcing bartenders to walk customers through a large, frequently updated inventory one by one. BeerMe reimagines the menu as a live, interactive layer for guests, while giving venue management a way to control inventory in real time.
Decision paralysis at scale: with 130+ rotating options, browsing a flat menu overwhelmed guestsPeak-hour service bottlenecks: personal guidance from bartenders drove satisfaction but severely slowed service during busy shiftsVenue managers need an ultra-simple admin interface they can update from a phone in seconds, with no downtime02 – Core Product Strategy
1Instant QR-to-PWA Onboarding
ProblemAsking guests to download an app or fill out a sign-up form at the table killed interest before they even saw the menu.
DecisionI built a fast Progressive Web App that opens straight from a QR code on the table, no install needed.
ReasoningZero friction at the first touch: the menu has to be on screen before the guest loses patience.
Impact
Guests go from scanning the code to browsing live inventory in seconds, on any phone.
2Interactive "Find Me a Beer" WizardProblemScrolling through 130+ beer names and brewing terms was overwhelming if you weren't already a beer person.
DecisionI built a quick 3-question flow that asks about flavor, intensity and style instead of jargon.
ReasoningIt meant someone who'd never heard of an IPA could still get a recommendation they'd actually enjoy.
Impact
Turns an overwhelming menu into 3 suggestions guests can pick from in seconds.
3Actionable Venue AnalyticsProblemVenue managers had no real way to know which beers guests actually liked, so reordering and menu decisions were mostly guesswork.
DecisionI added a stats view to the admin dashboard that separates what's ordered most, what's actually finished (not just poured), and what gets rated highest.
ReasoningThose three don't always point to the same beer, and that gap is exactly what managers needed to see.
Impact
Gives managers real numbers to base purchasing and menu decisions on, instead of going by memory.
03 – Results & Value
~6–8 weeks
Time to launch
27
Registered users
~52% (14 of 27)
Returning users
BeerMe was live on the bar's real menu with real guests. Early numbers are small, but the returning rate showed guests found it useful enough to use again on their next visit.
04 – Reflection
The biggest lesson from BeerMe was that the best opportunities are sometimes right where you already are. What started as a simple idea to digitize a menu turned into something that changed how guests order and how the bar tracks what's actually working. Building for a busy, real bar instead of a clean demo taught me to care more about removing friction than adding features.
Product Strategy · UX Research · Interaction Design · High-Fidelity UI · Rapid Prototyping · Front-end Implementation · Lovable · Supabase · AI-assisted Development
Read more about my experience
I'm Tom, a Product Designer who turns complex workflows into calm, intuitive products.I design, prototype, and build digital tools that solve real operational problems.
See selected works
Turning a single bar's scheduling chaos into a multi-tenant operations platform for hospitality teams.
01 – Background & Challenge
Shiftr started as a fix for one bar I worked at: schedules built in Excel, published as WhatsApp screenshots, and patched together in chat threads until nobody was sure who was actually working. The moment a second venue asked to use it, the problem changed shape — this was never a scheduling calendar, it was an operations platform that had to hold many independent businesses without ever letting one venue see another's data.
Managers don't need a calendar, they need a decision aid that holds availability, role eligibility and last-minute changesFloor staff use the product standing up, one-handed, in a loud room, for ninety seconds at a timeEvery venue thinks its rules are unique: role names, shift patterns and deadlines differed at every bar02 – Core Product Strategy
1Organisation as the Core Unit of the ProductProblemThe first version assumed one venue. Adding a second exposed that staff, roles, shifts and settings all had no owner — and that any leak between businesses would be fatal to trust.
DecisionRebuilt the product around an organisation entity: every user belongs to one or more organisations, an explicit org-selection step sets context, and all data access is scoped at the database layer.
ReasoningTenancy is a structural property, not a filter. Enforcing it beneath the UI means no future screen can accidentally expose another venue's roster.
Impact
Shiftr became a platform rather than a tool: new venues onboard independently, and multi-venue staff switch context in a single tap.
2Invite-Based Onboarding Instead of Open Sign-UpProblemAnyone could create an account and land nowhere — no venue, no role, no context — while managers had no reliable way to get their actual team into the system.
DecisionDesigned a tokenised invite flow: the link survives the whole sign-in or sign-up journey, and the new member lands inside the right organisation with the right role attached.
ReasoningIn workforce software, membership is granted by the business, not claimed by the user. Preserving the invite removes the most common drop-off point.
Impact
Team setup collapsed into one link, and no user ever reaches an empty, contextless state.
3Venue-Defined Roles and a Separate Permission LayerProblemHard-coded roles like 'Bartender' broke immediately, and tying job title to system capability meant scheduling rights also changed who could work which shift.
DecisionSplit the model in two: venues define their own roles for scheduling eligibility, while a separate permissions surface controls what each member can do in the system.
Reasoning'What you're qualified to work' and 'what you're allowed to change' are different questions. Separating them avoids over-granting access.
Impact
Every venue models its own team accurately, and permission changes became a deliberate, auditable act.
03 – Results & Value
Multi-venue
Independent organisations, fully isolated
1 link
Invite-based team onboarding
0 chasing
Automated availability reminders
The coordination work that used to live in a manager's head and a group chat now lives in the product. Venues onboard themselves, staff handle their own availability and swaps, and managers step in only where judgement is genuinely required.
04 – Reflection
Designing a multi-tenant product taught me that flexibility and opinion aren't opposites — they're a sequencing problem. Every venue's rules were unique, but only at the edges; underneath, the workflow was the same. I also learned how much of operational software is really permission design: who can change what is a product decision with real consequences.
I'm Tom - a Product Designer, UX/UI designer, and maker based in Israel. I sit at the intersection of visual craft, product thinking, and full-stack execution.
I believe good software should be quiet, focused, and respectful of user attention. Whether designing complex SaaS surfaces or refining mobile interactions, my goal is to eliminate friction and make complex workflows feel effortless.
I don't stop at static Figma prototypes. Using a modern full-stack workflow (Lovable, Supabase, React), I build and deploy working products into active real-world environments — from live venue engagement tools to internal shift operations platforms.
Outside the screen, I play chess, play the guitar and enjoy exploring the rich world of craft beers.
Career Timeline
2023 — Present
Product Designer & SaaS Maker
Independent Practice
Designing, building, and deploying web applications and SaaS tools - including BeerMe (a live venue tracking app) and custom shift management platforms.
2021 — 2023
UX/UI & Graphic Designer
Hila Advertisement & Co.
Crafted visual identities, digital marketing surfaces, and user interfaces for diverse commercial clients.
2017 — 2021
Freelance Graphic & Web Designer (Part-Time)
Freelance
Designed websites and brand identities for small local businesses, part-time alongside unrelated full-time work.
2010 — 2017
Editor & Product Designer
Photography Studio
Edited photography and designed client-facing products and marketing materials, alongside partial operations and coordination work.
Currently open to full-time Product Design roles, UX/UI positions, and high-impact product work. Based in Israel - available for on-site or remote roles.
DirectFastest way to reach me. Typically responding within a few hours.
Send A Message
A concept redesign of Israel's second-largest health fund member experience, focused on clarity, trust, and faster task completion.
01 – Background & Challenge
Healthcare member portals are notorious for dense menus, duplicate paths, and jargon. Meuhedet members often gave up and called the service center instead, creating cost and frustration for everyone.
Members think in tasks ("book an appointment") not in organizational categoriesTrust is the primary currency in health interfaces; clarity beats densityOlder users need larger touch targets but the same information hierarchy02 – Core Product Strategy
1Task-Based Home ScreenProblemThe existing homepage mirrored the organization chart, not member needs.
DecisionRedesigned the home around the five tasks members do most often, with clear entry points.
ReasoningThis reduced cognitive load and gave members a confident starting point every visit.
Impact
Task-start rate increased 28% in usability testing.
2Plain-Language LabelsProblemTerms like "authorization" and "referral" confused members.
DecisionRewrote navigation labels in everyday Hebrew and added short explanations.
ReasoningMembers understood what each action meant before clicking.
Impact
Support calls for navigation help dropped by a third.
3Unified Appointment ViewProblemAppointments were scattered across multiple sections.
DecisionCreated a single timeline of upcoming and past appointments.
ReasoningOne place to look means members stop hunting and start acting.
Impact
Members found appointment details 40% faster.
03 – Results & Value
+28%
Task-start rate
-33%
Navigation support calls
-40%
Time to find appointment
The redesign showed that a calmer, task-centered interface can rebuild trust in a complex healthcare system without sacrificing depth. Figures come from usability testing of the concept.
04 – Reflection
Healthcare UX is less about delight and more about removing fear — every extra click feels like a risk to users. Designing for health meant choosing clarity and calm over novelty, and speaking the member's language instead of the organization's.
I don't separate design from making. I sketch ideas directly through prototypes, test them quickly, and iterate until the product feels sharp. The result is software that helps people move with less friction and more confidence.
Read more about my practice
Case Study 01
Helping craft beer bars turn menus into interactive experiences.
01 – The Problem
While working at a renowned craft beer bar, staff managed over 130 rotating beers. Printed menus quickly became outdated as the selection changed, creating friction for guests and forcing bartenders to walk customers through a large, frequently updated inventory one by one. BeerMe reimagines the menu as a live, interactive layer for guests, while giving venue management a way to control inventory in real time.
02 – Key Decisions
ProblemAsking guests to download an app or fill out a sign-up form at the table killed interest before they even saw the menu.
DecisionI built a fast Progressive Web App that opens straight from a QR code on the table, no install needed.
ReasoningZero friction at the first touch: the menu has to be on screen before the guest loses patience.
Impact
Guests go from scanning the code to browsing live inventory in seconds, on any phone.
ProblemScrolling through 130+ beer names and brewing terms was overwhelming if you weren't already a beer person.
DecisionI built a quick 3-question flow that asks about flavor, intensity and style instead of jargon.
ReasoningIt meant someone who'd never heard of an IPA could still get a recommendation they'd actually enjoy.
Impact
Turns an overwhelming menu into 3 suggestions guests can pick from in seconds.
ProblemVenue managers had no real way to know which beers guests actually liked, so reordering and menu decisions were mostly guesswork.
DecisionI added a stats view to the admin dashboard that separates what's ordered most, what's actually finished (not just poured), and what gets rated highest.
ReasoningThose three don't always point to the same beer, and that gap is exactly what managers needed to see.
Impact
Gives managers real numbers to base purchasing and menu decisions on, instead of going by memory.
03 – Outcomes & Impact
~6–8 weeks
Time to launchEst.
27
Registered users
~52% (14 of 27)
Returning usersEst.
BeerMe was live on the bar's real menu with real guests. Early numbers are small, but the returning rate showed guests found it useful enough to use again on their next visit.
04 – Lessons Learned
The biggest lesson from BeerMe was that the best opportunities are sometimes right where you already are. What started as a simple idea to digitize a menu turned into something that changed how guests order and how the bar tracks what's actually working. Building for a busy, real bar instead of a clean demo taught me to care more about removing friction than adding features.
Product Strategy · UX Research · Interaction Design · High-Fidelity UI · Rapid Prototyping · Front-end Implementation · Lovable · Supabase · AI-assisted Development
Read more about my experience
I'm Tom, a Product Designer who turns complex workflows into calm, intuitive products. I design, prototype, and build digital tools that solve real operational problems.
See selected works
Case Study 02
Turning a single bar's scheduling chaos into a multi-tenant operations platform for hospitality teams.
01 – The Problem
Shiftr started as a fix for one bar I worked at: schedules built in Excel, published as WhatsApp screenshots, and patched together in chat threads until nobody was sure who was actually working. The moment a second venue asked to use it, the problem changed shape — this was never a scheduling calendar, it was an operations platform that had to hold many independent businesses without ever letting one venue see another's data.
02 – Key Decisions
ProblemThe first version assumed one venue. Adding a second exposed that staff, roles, shifts and settings all had no owner — and that any leak between businesses would be fatal to trust.
DecisionRebuilt the product around an organisation entity: every user belongs to one or more organisations, an explicit org-selection step sets context, and all data access is scoped at the database layer.
ReasoningTenancy is a structural property, not a filter. Enforcing it beneath the UI means no future screen can accidentally expose another venue's roster.
Impact
Shiftr became a platform rather than a tool: new venues onboard independently, and multi-venue staff switch context in a single tap.
ProblemAnyone could create an account and land nowhere — no venue, no role, no context — while managers had no reliable way to get their actual team into the system.
DecisionDesigned a tokenised invite flow: the link survives the whole sign-in or sign-up journey, and the new member lands inside the right organisation with the right role attached.
ReasoningIn workforce software, membership is granted by the business, not claimed by the user. Preserving the invite removes the most common drop-off point.
Impact
Team setup collapsed into one link, and no user ever reaches an empty, contextless state.
ProblemHard-coded roles like 'Bartender' broke immediately, and tying job title to system capability meant scheduling rights also changed who could work which shift.
DecisionSplit the model in two: venues define their own roles for scheduling eligibility, while a separate permissions surface controls what each member can do in the system.
Reasoning'What you're qualified to work' and 'what you're allowed to change' are different questions. Separating them avoids over-granting access.
Impact
Every venue models its own team accurately, and permission changes became a deliberate, auditable act.
03 – Outcomes & Impact
Multi-venue
Independent organisations, fully isolated
1 link
Invite-based team onboarding
0 chasing
Automated availability reminders
The coordination work that used to live in a manager's head and a group chat now lives in the product. Venues onboard themselves, staff handle their own availability and swaps, and managers step in only where judgement is genuinely required.
04 – Lessons Learned
Designing a multi-tenant product taught me that flexibility and opinion aren't opposites — they're a sequencing problem. Every venue's rules were unique, but only at the edges; underneath, the workflow was the same. I also learned how much of operational software is really permission design: who can change what is a product decision with real consequences.
I'm Tom - a Product Designer, UX/UI designer, and maker based in Israel. I sit at the intersection of visual craft, product thinking, and full-stack execution.
I believe good software should be quiet, focused, and respectful of user attention. Whether designing complex SaaS surfaces or refining mobile interactions, my goal is to eliminate friction and make complex workflows feel effortless.
I don't stop at static Figma prototypes. Using a modern full-stack workflow (Lovable, Supabase, React), I build and deploy working products into active real-world environments — from live venue engagement tools to internal shift operations platforms.
Outside the screen, I play chess, play the guitar and enjoy exploring the rich world of craft beers.
Career Timeline
2023 — Present
Product Designer & SaaS Maker
Independent Practice
Designing, building, and deploying web applications and SaaS tools - including BeerMe (a live venue tracking app) and custom shift management platforms.
2021 — 2023
UX/UI & Graphic Designer
Hila Advertisement & Co.
Crafted visual identities, digital marketing surfaces, and user interfaces for diverse commercial clients.
2017 — 2021
Freelance Graphic & Web Designer (Part-Time)
Freelance
Designed websites and brand identities for small local businesses, part-time alongside unrelated full-time work.
2010 — 2017
Editor & Product Designer
Photography Studio
Edited photography and designed client-facing products and marketing materials, alongside partial operations and coordination work.
Currently open to full-time Product Design roles, UX/UI positions, and high-impact product work. Based in Israel - available for on-site or remote roles.
DirectSend A Message
Case Study 03
A concept redesign of Israel's second-largest health fund member experience, focused on clarity, trust, and faster task completion.
01 – The Problem
Healthcare member portals are notorious for dense menus, duplicate paths, and jargon. Meuhedet members often gave up and called the service center instead, creating cost and frustration for everyone.
02 – Key Decisions
ProblemThe existing homepage mirrored the organization chart, not member needs.
DecisionRedesigned the home around the five tasks members do most often, with clear entry points.
ReasoningThis reduced cognitive load and gave members a confident starting point every visit.
Impact
Task-start rate increased 28% in usability testing.
ProblemTerms like "authorization" and "referral" confused members.
DecisionRewrote navigation labels in everyday Hebrew and added short explanations.
ReasoningMembers understood what each action meant before clicking.
Impact
Support calls for navigation help dropped by a third.
ProblemAppointments were scattered across multiple sections.
DecisionCreated a single timeline of upcoming and past appointments.
ReasoningOne place to look means members stop hunting and start acting.
Impact
Members found appointment details 40% faster.
03 – Outcomes & Impact
+28%
Task-start rate
-33%
Navigation support calls
-40%
Time to find appointment
The redesign showed that a calmer, task-centered interface can rebuild trust in a complex healthcare system without sacrificing depth. Figures come from usability testing of the concept.
04 – Lessons Learned
Healthcare UX is less about delight and more about removing fear — every extra click feels like a risk to users. Designing for health meant choosing clarity and calm over novelty, and speaking the member's language instead of the organization's.
