Web
Website creation
Modern websites, landing pages, and portfolio builds with stronger hierarchy, better responsiveness, and cleaner visual direction.
Portfolio
Selected work across
websites, interfaces,
and custom systems.
The portfolio mixes front-facing digital experience with more operational builds — because many projects need both: a sharper interface on the surface and stronger logic underneath.
Capability tracks
The strongest through-line is modern web creation, supported by more technical systems when needed.
UX
Custom interfaces
Purpose-built UI and UX work that makes the interaction feel clearer, lighter, and more intentional.
Ops
Workflow systems
Automation-backed tools, marketplace workflows, and structured systems designed around real operational use.
Local
Restaurant & local SEO
Conversion-focused websites for hospitality businesses, with schema.org markup, geo-meta strategy, and design that reflects the venue's identity.
Case studies
Four projects in depth.
Case study 01
Planet Wok — Restaurant website
A full-page, SEO-first website for an Asian all-you-can-eat restaurant in Quetigny (Dijon), built to rank locally, convert visitors into reservations, and reflect the premium ambience of the venue from the first scroll.
- React 19 + Vite
- Tailwind CSS v4
- Framer Motion
- Local SEO
- Schema.org / JSON-LD
- Dark luxury design
The brief
Planet Wok had no dedicated website — reservations came entirely from Google Maps and word of mouth. With 3,400+ Google reviews and a 4.5/5 rating, the business had strong social proof but nowhere to send visitors who wanted to see the menu, pricing, or atmosphere before booking.
The approach
Designed around a “Feu & Nuit” (Fire & Night) concept — deep charcoal backgrounds, burnt-gold accents, and Playfair Display headings — to evoke the premium evening dining experience. Full schema.org Restaurant + FAQPage markup, geo-meta tags, and a French-language keyword strategy were built in from the start, not bolted on. Framer Motion parallax and scroll-reveal animations add drama without sacrificing performance.
What was delivered
A single-page React app covering hero, buffet showcase, full menu, pricing tables (adult & children, lunch & dinner), curated Google reviews, an interactive map, and a contact section. Structured data makes the site immediately readable by Google, AI assistants, and local search aggregators — giving the restaurant visibility beyond a standard listing.
Case study 02
Visual recognition inventory app
A mobile-oriented inventory system designed to pair visual recognition with connected stock data so operators can identify items quickly and keep inventory changes visible across the workflow.
- Computer vision
- Inventory software
- Mobile UX
- API integration
The problem
Manual inventory handling is slow, error-prone, and difficult to monitor when item recognition and stock updates depend on separate tools or human memory.
The approach
Combine vision-assisted recognition, a mobile interface for operators, and backend connectivity so status changes travel through one coherent flow instead of fragmented spreadsheets and messages.
Why it matters
This type of build fits teams that need inventory visibility, lower manual overhead, and interfaces grounded in the speed of the warehouse or field environment.
Case study 03
Multi-marketplace listing engine
A publishing workflow for creating once and distributing across eBay, Leboncoin, Vinted, and Amazon while handling the differences that normally force teams back into manual repetition.
- Marketplace automation
- Catalog normalization
- Cross-platform workflows
- Operational tooling
The problem
Listing products on multiple marketplaces usually means duplicated effort, inconsistent data, and platform-specific fields that create friction every time the catalog changes.
The approach
Centralize product data, normalize reusable fields, then automate channel-specific publishing rules so the workflow scales without requiring teams to rebuild each listing by hand.
Why it matters
This work is relevant for sellers, operators, and platform-heavy businesses that need faster throughput, fewer posting mistakes, and cleaner control over their catalog.
Case study 04
Scraper bot & pricing intelligence stack
Structured data collection for monitoring competitor pricing, tracking market visibility, and feeding pricing research pipelines with reliable, repeatable data.
- Web scraping
- Pricing intelligence
- Data pipelines
- Competitive analysis
The problem
Pricing decisions made without current market data are slow and reactive. Manual research doesn't scale, and most off-the-shelf tools don't fit the specific data shape the business needs.
The approach
Build targeted scrapers for the relevant platforms, normalize the output into a consistent schema, and feed it into a lightweight dashboard or reporting layer the team can actually use.
Why it matters
This work fits businesses that compete on price, need faster market signals, or want to stop relying on guesswork when setting rates or reacting to competitor moves.
Start a project
Ready to build something that works as well as it looks?
Email Thomas with the project brief, goal, or redesign need. Direct email is the fastest way to get started.
thomas@talentdeck.me
Best contact: direct email