Loading...
Founder & CEO · Hyperse
As Co-founder, CEO & CTO of Hyperse, I set the technical and product bet: Vendure + Next.js + GraphQL, a licensed plugin platform, and live multi-market brands I operate myself. Clients inherit the same architecture — not a one-off build.
Lead the platform · Operate the proof · Productize the delivery
Storefront
Next.js · GraphQL · i18n
Commerce core
Vendure · Admin · GraphQL
Hyperse Hub
@hyperse-hub · 60+ plugins
Licensed capability domains brands assemble — not a theme fork.
Open source
100+ packages on npm under @hyperse/* — hyperse-io
Operated proof
kezvor multi-market · sister brands · AlphaIO
I operate these surfaces — open them before you trust a slide.
Leadership span plus operated outcomes — no invented GMV. Delivery tempos match published Hyperse engagement packages.
Sprint / diagnostic tempos are the fixed packages I sell (Plugin Sprint · Commerce Audit) — engineering delivery, not revenue guarantees.
Selected systems I led — owned brands and platform bets you can open in a browser.
Problem
Optical isn’t another fashion SKU — lens configuration, trust content, multi-market ops, and lifecycle messaging break generic themes and agency handoffs.
Decision
I bet on a pluginized Vendure core + Next.js storefronts, with Help as GEO engine and client/server tracking — then operated US / DE / UK myself instead of outsourcing the proof.
Result
Flagship result: 3 live markets (US · DE · UK) on one commerce core; sister brands reuse pay/admin/catalog without forking the stack — verify at kezvor.com / .de / .co.uk.
Problem
Treating each brand as a separate project creates drift until payments, catalog, and admin cannot be maintained.
Decision
Leadership rule: capabilities sink into the plugin platform; brands only vary theme, offer, and merchandising.
Result
4+ brand fronts on the same family (Kezvor + FOGLAX + KZFOO and siblings) — launches become assembly on the hub, not a rewrite of commerce.
Problem
Agencies sell hours. A CEO needs a public face that matches repeatable capability — plugins, OSS, and operated proof.
Decision
I positioned Hyperse around licensed @hyperse-hub domains plus hyperse-io standards, with US LLC + Shanghai ops for cross-border credibility.
Result
A clickable company surface (hyperse.net) aligned with how we actually deliver — Community / Pro / Enterprise paths for teams building on the framework.
Problem
Most API portals fake packaging: shared rate limits, demos that diverge from production, muddy compliance.
Decision
Sell by featureCode with subscribe/preview states; per-feature quotas; one Shop GraphQL contract — schema-driven product thinking.
Result
A credible pattern for monetizing datasets and metered APIs — research data, not trading advice.
I am a hands-on CEO: I still own architecture when the bet is existential, and I organize delivery so capability compounds into a platform — not a pile of projects. Hyperse exists to help brands own their commerce stack; Kezvor and sister brands exist to prove that stack under real operations.
Own the platform, don’t rent a theme. Headless Vendure + Next.js keeps margin, multi-brand reuse, and engineering control with the brand — not the marketplace.
As Engineering Manager at Tongcheng Fintech (2015–2025), I led up to 40+ across FE/BE/Mobile/UX/QA — full-chain delivery, not a front-end silo. At Hyperse I keep a smaller core and productize what used to be heroic one-offs.
I run multi-market optical D2C (Kezvor and sister brands) on the same platform I sell. If tracking, lens ops, or multi-channel break, I feel it in production — not in a PDF.
One-off builds become @hyperse-hub domains — with Admin shell and marketplace licensing. Reuse is a leadership outcome, not a hope.
Full-time Co-founder, CEO & CTO — set the company bet on owned headless commerce, ship the licensed plugin platform, operate multi-market brands as proof, and keep open-source tooling under hyperse-io as the engineering standard.
Earlier
Scaled Hyperse from early ecosystem work while still leading platform engineering at Tongcheng — dual track that forced productization over pure custom delivery.
Platform leadership across Web + mobile surfaces: architecture governance, Hybrid/Web/Node unity, and engineering standards so Android/iOS/FE ships on one cadence.
Tongcheng Fintech · Engineering Manager — tech planning for up to 40+ FE/BE/Mobile/UX/QA; built CI/CD mid-platforms, GraphQL multi-project governance, and quality systems that outlived individual apps.
IC → lead path through CSSoft, Shanda, and product companies — .NET/front-end into SPA architecture and modular platforms.
Scaled Hyperse from early ecosystem work while still leading platform engineering at Tongcheng — dual track that forced productization over pure custom delivery.
Platform leadership across Web + mobile surfaces: architecture governance, Hybrid/Web/Node unity, and engineering standards so Android/iOS/FE ships on one cadence.
Tongcheng Fintech · Engineering Manager — tech planning for up to 40+ FE/BE/Mobile/UX/QA; built CI/CD mid-platforms, GraphQL multi-project governance, and quality systems that outlived individual apps.
IC → lead path through CSSoft, Shanda, and product companies — .NET/front-end into SPA architecture and modular platforms.
Leadership is not a title — it is what compounds after one project ends.
Choose owned headless boundaries — storefront, checkout, admin — that a brand can still run in year two.
Turn delivery into @hyperse-hub plugins, Admin shell, GraphQL contracts — and the same discipline on mobile/Web: shared pipelines, not per-app folklore.
Hold the line on tracking accuracy, Help/GEO, affiliate/membership, feeds & lifecycle — growth is a platform concern, not a campaign afterthought.
The system bet behind Hyperse — and the hard problems I refuse to outsource as “someone else’s ticket.”
Next.js + GraphQL + i18n — brand faces that stay fast and ownable.
Schema-driven Shop/Admin GraphQL — one contract, many clients.
Vendure core with @hyperse-hub domains — capability assembly per brand/channel.
Tracking, Help/GEO, feeds, pay/ship — wired into the platform, not pasted later.
I productize delivery into installable domains — visitors get the map; engineers get the package catalog on GitHub.
Platform
Admin shell, licensing, and identity — the runtime every brand front assembles on.
Commerce & ops
From catalog and pay through ship, growth, and support — capabilities sink into plugins so brands don’t fork the core.
Vertical depth
Optical flows that generic commerce never ships — the reason Kezvor isn’t a fashion theme.
These are why generic themes and agency handoffs fail in production.
Lens flows, trust content, and vertical plugins (Hub Glass) — not fashion SKUs with extra fields.
Data, permissions, and config isolation so multi-brand reuse does not become shared-state chaos.
Client + server pipelines with typed contracts — so ads and lifecycle email stop lying.
Featured system
Open-source typed tracking for GA4 · Meta · Klaviyo. A CEO decision: measurement is a platform concern, not a GTM afterthought.
POC → live can be weeks when the platform already exists — the constraint is decision quality and vertical complexity, not scaffolding a theme.
Commerce platform first — with mobile delivery and an open @hyperse/* toolchain as proof I compound engineering assets, not only storefronts.
Vendure
Next.js
GraphQL
TypeScript
NestJS
@hyperse-hub
Android
iOS
Hybrid / WebView
@hyperse/*
@hyperse/track
@hyperse/hps
CI/CD · Sentry
100+ packages under @hyperse — the DX standard behind Hyperse delivery. Flagships below; full inventory on npm.
Typed tracking and logging so ads and lifecycle stop lying.
Rspack-era HPS and a type-safe Wizard core for internal tools.
Click-to-source inspector and type-safe Next config patterns.
ICU translator and shared ESLint — the floor every repo inherits.
Hub domains, hard problems, and @hyperse/track live in System.
No stock avatars. Verify the work in a browser, then decide if founder-led delivery is what you want.
You’ll get founder-level architecture and the person who runs production brands — not a revolving door of juniors learning on your milestone.
Send stack, timeline, constraints, and what success means in year two. I reply with a concrete approach — usually a fixed package or a short paid diagnostic. Casual ‘just looking’ shopping is fine on hyperse.net.