Wert product overview

Developing an Embedded Checkout, End to End

1.1M+ registered users in 2025
1.2M+ successful orders in 2025

Wert is an embeddable checkout that lets partners' users buy and sell digital assets inside their apps and websites. The hard part is making KYC, onboarding, and compliance feel straightforward, while supporting hundreds of partner setups that all behave differently.

I built Wert as the founding designer between 2020–2023. After occasionally consulting the team on new features, in January 2025 the company brought me back as Design & Product Lead. The product was scaling fast through new partnerships and needed tighter ownership to keep the experience consistent.

2025–2026 Wert Design & Product Lead

I led a cross-functional squad (backend, two frontend engineers, QA), set design direction across the widget and internal tools, and owned key product decisions across checkout, KYC, and partner flows. A major part of my work was identifying and fixing high-impact conversion issues that had emerged as the product scaled.

In startups, internal tools are rarely a priority. You can brute-force operations with manual work until volume makes it painful. As Wert scaled, the Admin Panel became a bottleneck. Support, Sales, and Compliance used it daily for lookups, order management, and partner configs, but it didn't reliably support those jobs. On top of that, the platform was mid-migration to a new backoffice layer, so the underlying services were changing while teams still depended on the old ones.

The goal was to keep Support fast and reliable as volume grew, without waiting for the migration to finish.

Approach

JTBD first. When things look broken, the risk is trying to fix everything at once. I mapped the actual jobs teams did daily and ranked them by frequency and risk. That told us what to fix first. What's slowing the team down most right now?

Quick wins built trust. We fixed the most painful flows first, the ones that forced teams to switch tools and work around bugs. That proved the work was worth it and gave the business confidence to support further improvements.

Progressive delivery. With the backend mid-migration, not everything was available at once. Shipped in small releases as new endpoints came online, coordinating with backend on what was feasible in each rollout.

Patterns and guardrails. I set up reusable UX patterns and templates for screens, forms, and UI states. The goal was to let the team ship new and updated flows without pulling Design into every ticket.

Documentation. Documented workflows and expected outcomes so the team had a shared reference instead of tribal knowledge. When things change fast, written clarity is what keeps teams aligned.

Impact

  • New jobs covered by Admin Panel reduced tool-switching and manual workarounds.
  • Time on task dropped by 30–50% across core flows due to improved UX.
  • Design patterns and templates sped up future releases. The team could ship new features without reinventing the UI each time.
  • Kept operations stable throughout the migration, which was the hardest part.

↓ 30–50% time on task for core support actions

Wert Admin Panel dashboard

Document verification has a high failure rate across the industry — according to SumSub, up to 50% of KYC checks require resubmission (poor lighting, blurry photos, expired documents, mismatched data). Wert was no exception. Verification for larger purchases can take minutes or hours, so few stay on screen. Users rely on follow-up emails to complete the process.

After a necessary major security update, we changed how the widget could be triggered (the details are sensitive). The resubmission flow broke: users could no longer return to their verification directly from emails and had to restart the entire purchase from the partner's website. Re-verification drop-off rose from 52% to 64%. High-value user KYC recovery fell from 15% to 9%.

Approach

Map the full journey, not just the break. After a UX audit, the problems went beyond broken links. Emails didn't reflect the case or context, had no design, and were easily replicated by fraudsters. Support couldn't help users without routing them through partner sites. No tracking to understand where users dropped.

Reframe the problem. Resubmission isn't a checkout retry. It's a separate user journey with different context, timing, and trust. Users come back hours or days later, often from a different device, with little context of where they left off. I proposed separating resubmission from the purchase flow entirely with a standalone widget that works without a partner session.

Redesign every touchpoint. Transactional emails redesigned for clarity, branding, and trust. Direct resubmission link for Support. So, we no longer send users to partner sites for KYC issues. Amplitude tracking across the new flow so we could see what was actually happening.

Drive it across squads. This wasn't on my squad's roadmap. I saw the drop-off numbers, made the case for urgency, set up the work with a frontend lead outside my squad, and we shipped everything within a month.

Impact

  • New standalone resubmission flow. Users could now return and complete verification directly, from any channel.
  • Redesigned emails with clear context, step-by-step guidance, and trustworthy design.
  • Re-verification drop-off: 52% before → 64% after security update → 27% after release.
  • High-value user KYC recovery (failed → completed): ~15% before → 9% after security update → ~17% after release and held steady at ~18% median over the following quarter.

↓ Re-verification drop-off: 52% → 27%

↑ High-value users KYC recovery (failed → completed): 15.8% → 17.5%

↑ New median 18%

Wert emails — before redesign Wert emails — after redesign

Mobile was 75%+ of traffic and converted better. But we were still running a desktop-first layout I had designed years earlier. Built to deliver a full experience across all screen sizes. It made sense at the time. It didn't anymore. Supporting it meant extra component states, more edge cases, and added complexity for a product that needed to stay small and fast. The layout was costing more than it was giving back.

Data backed it up. I led the decision to deprecate the desktop layout entirely and shift to mobile-first. Alongside that, I introduced self-contained modular components: payment method, card selection, wallet address. Each independently designable and buildable. Less coupling, less redundant work per feature, ~2× design and frontend shipping speed.

~2× design and frontend shipping speed

Buy Bitcoin
You pay
1,200 USD
You get
0.05579674 BTC
Purchase details

For me, this work is a snapshot of how I operate: moving from user pain to metrics, aligning priorities, and shipping changes that hold up in production. I ran the squad with distributed ownership, with team members owning features end to end while I kept a high bar for product thinking and craft.

I had the pleasure of working closely with Gamzat at Wert. He has been a core part of Wert since its foundation. As the lead designer, he shaped the user experience for our payment module and partner accounts and managed our design department effectively. His clear and consistent communication with me and the broader management team has been invaluable. I've always been impressed with how swiftly and lucidly he anticipates future design trajectories and potential challenges, guiding us towards the best solutions.

George Basiladze, Founder at Wert

Buzzbike bike close-up
Buzzbike app map view

Improving Rider and Fleet Team Experience

3.5k+ active paying subs in 2023
4.4 App Store rating

Buzzbike was a London-based bike subscription service offering monthly plans with delivery and on-demand support. The main challenge was translating heavy, real-world operations (fleet management, repairs, deliveries, and the rider app) into a product experience that felt simple and fair. In late 2024, rising operational costs made the model unsustainable, and the company shut down.

I joined Buzzbike as Lead Designer. The product already had a strong visual identity. I helped sharpen and systematise it. Alongside that, I pushed for stronger UX thinking: challenging briefs, clarifying flows, raising execution quality. Over time I helped ship clearer rider experiences and internal tools used daily by ops teams.

2023–2024 Buzzbike Lead Designer
Onboarding — on-demand repairs
Onboarding — track stats

Designed the Buzzbike Ops app used by mechanics & drivers to manage fleet operations day to day

Buzzbike's operations relied on constant deliveries, repairs, and warehouse work, but the workflow lacked clear tools and transparency. Leadership initiated an internal Ops app to coordinate work and make it visible.

I worked with the founders and ops leads who scoped the initiative and defined requirements, then partnered directly with mechanics and drivers to validate real workflows. I designed and tested the Ops app through prototypes and test builds, iterating on field feedback to simplify flows and surface the right context.

Driver delivery screen

Reduced avoidable mechanic callouts by redesigning how repairs are requested in-app

Buzzbike had unlimited free mechanic callouts and a "straight to help flow", which led to many unnecessary requests for minor issues. The business decided to cap free bookings to a fixed number per year, with additional callouts becoming paid. My task was to translate that into a flow that stayed supportive rather than feeling like a paywall. I talked with the mechanics to understand which issues could be resolved without a visit, then redesigned how riders report problems by adding lightweight triage through clearer issue selection and follow-up questions, along with a simple decision checkpoint before booking. This led to a clear drop in unnecessary callouts while keeping mechanic support accessible.

Buzzbike mechanic at work
Support screen
Booking profile redesign

Improved conversion critical journeys including sign-up and cancellation through iterative UX fixes

I contributed across the rider-facing product, including sign-up, onboarding, cancellation, and ongoing updates to the website and mobile app. The team leaned heavily on A/B testing to guide decisions, but with relatively low traffic I was careful not to treat results as definitive. I pushed for a more balanced approach, using experiments as signals while relying on product judgment and UX reasoning to help the team decide what was worth shipping and scaling.

Clarifying Core Workflows for an AI Automation Platform

Scade is an AI automation startup focused on helping teams build and deploy AI workflows without writing code. I led product design efforts to clarify core workflows and improve user activation across the platform.

improvement in onboarding completion
40% reduction in time to first workflow

I joined to untangle a complex product. The platform had powerful capabilities but users struggled to understand what they could do and where to start. My focus was making the core value proposition clear through the interface.

Working closely with the founding team, I redesigned the workflow editor, simplified the onboarding experience, and established patterns that scaled across different automation types.

Scade • 2024–2025 Product Design Lead

Selected Impact

* figures are shown as relative changes; some are exact, others are quarterly medians

AI tools are only useful if people can understand what they do and trust the outputs. My job was to make the complexity feel approachable without dumbing it down

The original onboarding dropped users into the workflow editor with no context. Most users churned before creating their first automation because they didn't understand what was possible or where to begin.

I redesigned onboarding around use cases rather than features. Users now start by selecting what they want to accomplish, and the system guides them through a relevant template that demonstrates the core concepts.

This shift from "here's what the tool does" to "here's what you can do" doubled the completion rate and got users to their first success faster.

↑ 2× improvement in onboarding completion

The workflow editor was powerful but overwhelming. Users couldn't easily see the flow of data between steps, and the node-based interface required too much manual configuration.

I simplified the visual model, added smart defaults, and introduced inline previews that let users see outputs at each step. The result was a faster path to working automations with less trial and error.

↓ 40% reduction in time to first workflow

As the product expanded to support more automation types, we needed patterns that could scale. I established a component system and interaction patterns that worked across different contexts while maintaining consistency.

This foundation made it faster to ship new features and gave the engineering team clear guidance on how new capabilities should look and behave.

↑ Established scalable design patterns

Building Payment Module for Web3 from scratch

This case is from my time as Founding Designer at Wert, written a few years ago. The work and reasoning are real, but I'd frame and prioritise the story differently today, based on how my role and perspective have evolved. Also, honestly, it's a bit boring to read. Maybe one day I'll replace it with a short AI summary.

For my recent work at Wert, see the 2025, Design & Product Lead Case →

PS: Wert is an embeddable checkout that lets partners' users buy and sell digital assets inside their apps and websites. The hard part is making KYC, onboarding, and compliance feel straightforward, while supporting hundreds of partner setups that all behave differently.

2020–2023 Wert Founding Designer
Saudi Arabia pitch deck design

Designing Enterprise & Fintech in a Leading Design Agency

Clay is a San Francisco UI/UX and branding agency that designs digital products, websites, and design systems for startups and major tech brands. The challenge was delivering polished work fast while constantly switching between clients and styles, often without full visibility into real constraints, data, and what happens after handoff.

At Clay, I worked primarily on fintech and enterprise products. The work included two projects: a redesign of key transactional flows for the DSX.UK crypto exchange, and enterprise analytics platform for Brightfield.

2019–2020 Clay Senior UX/UI Designer

I joined Clay to strengthen visual craft and raise my execution bar. I already had strong product and UX fundamentals, but I wanted to learn in a team where craft is the standard.

Selected Projects

Brightfield — designing a powerful workforce analytics platform

Brightfield is a workforce analytics SaaS company. Their platform, TDX (Talent Data Exchange), uses data and automation to help enterprises benchmark roles and rates, track trends, and find cost and risk opportunities across contingent labour and SOW spend.

I worked as the Senior UX/UI Designer alongside a UX Director and Art Director, defining the design language and applying it across core product flows. That included setting the visual system and interaction patterns for a dense, data-heavy interface, then designing key screens and workflows so complex information stayed consistent, readable, and usable. It became the foundation for the live product's UX and visual language.

DSX Crypto Exchange — redesigning the core money-making flows

Discontinued

Clay led a full redesign of DSX, a UK crypto exchange. My focus was the Deposits area and related entry points, where friction and unclear fee visibility were causing confusion.

What I delivered:

  • Reworked deposit entry points and the deposit flow structure
  • Added clearer guidance around steps, limits, and expected outcomes
  • Made fees and method details more visible to reduce "hidden charges" perception
  • Improved consistency of UI patterns and primary actions within the Deposits section

The redesigned Deposits experience shipped to a test cohort. The client reported fewer deposit-related support tickets and an increase in weekly deposit volume during the test period.

teXet rugged phone

Crafting Smartphones & Strategy

500k+ smartphone units sold (X-basic)
4.0 avg rating on Yandex Market

Texet (GSM / Smartphone Department) was a mass-market consumer electronics brand sold through retail chains in Russia/CIS: phones, smartphones, tablets, e-readers, and more. The market reality was brutal. Low margins, intense price competition, uneven supplier quality (hardware + software), and unforgiving customers.

I was lucky to join teXet's new smartphone department in 2012, with a clear goal: run a pilot year, then secure reliable partners and scale into a fully operational department shipping well into six figures annually. Flamefox was a smaller, EU registered brand focused on smartphones + feature phones.

2012–2017 teXet / Flamefox PM / Industrial Designer

I'm highlighting this project because it's one of the clearest examples of how design thinking and UCD can drive outsized results.

Strategy

Rather than importing generic Chinese devices, we used the rise of low-cost Android platforms to build a durable position in the low-end segment. My team made a deliberate bet: in a commodity market, specs don't build loyalty, trust does. So we shaped our proposition around what competitors ignored: a clear day-one experience (improving Android UX and localisation) and support that helped people learn the device. That's what helped us win retail distribution.

So we did the following:

  1. Built phones for clear target groups (not "one model for everyone")
    • Elderly: large icons, simplified UI, SOS + family shortcut.
    • Rural / active: rugged devices built for durability.
    • Main audience: better UX + localisation, smart spec trade-offs (e.g., IPS over memory).
    • Mid segment: small higher-spec batches to test positioning before scaling.
  2. Made support part of the product
    • Proactive support team + support page/contact form + dedicated elderly hotline.
    • Tracked impact via in-store questionnaires.
  3. Enabled power users. Shared firmware/tools for flashing, experimentation, and occasional self-repair.

1.2M+ smartphones and feature phones sold over three years

teXet team photo
Flamefox DRIVE specifications

This combination of budget hardware with a reliable experience (strong UX, proper localisation, and real customer support) worked especially well in low-income markets, where trust and value matter more than top specs. We scaled fast: in 2013 the GSM Department was named Best Department of the Year and shipped 514,000 units (290,000 feature phones, 224,000 smartphones), doubling unit volume vs 2012 and tripling revenue; X-basic alone reached ~300,000 units sold.

Design Systems at Scale

DateMarch 2025
TopicProcess

Thoughts on maintaining design consistency across multiple products and teams.

On Founding Design Roles

DateNovember 2024
TopicThoughts

What I've learned about being the first designer at a startup.

Figma Workflow Tips

DateAugust 2024
TopicTools

A collection of Figma techniques I use daily.