Client
Joselito, Spain's ultra-premium Jamón Ibérico producer
My role
Product Lead, concept, UX architecture & prototype (end to end)
Team
Group of 6, IE MiM
Presented to
CEO of Joselito, Jose Gomez
The problem
A brand built on expertise, that went silent online
Joselito has been producing Jamón Ibérico de Bellota for six generations. It does not need mass advertising, Michelin-starred chefs across the globe carry it, and Ferran Adrià has called it the best ham in the world. The product has never been in question.
What was in question was the gap between Joselito's in-store experience and its digital one. In a physical boutique, a cortador can guide a customer through every product, the cut, the pairing, the right wine, how to store it overnight. That depth of service simply did not exist online, and it was quietly costing Joselito customers, professional partners, and international reach.
More than 70% of gourmet shoppers begin their purchase journey online, yet Joselito's website offered no guidance on how to choose, slice, or serve its products. The knowledge that defined the brand was locked in physical stores and expert staff.
What the brand was experiencing
- Customers unsure how to choose between products
- No slicing, storage or serving guidance → hesitation
- Website didn't match in-store prestige or warmth
- Chefs and distributors relying on manual email chains
- Certifications and training materials scattered or absent
- Hard to scale internationally without a digital backbone
What the strategy needed to deliver
- Replicate the maître d'hôtel experience, digitally
- Educate and guide without sounding like a FAQ bot
- Speak in Joselito's voice, warm, refined, never transactional
- Automate B2B partner operations at scale
- Serve Chinese and international markets multilingually
- Drive conversion without compromising luxury positioning
The solution
Joselito AI, Your Digital Waiter
An intelligent AI assistant built into Joselito's website and connected to WhatsApp Business and WeChat. It behaves the way a sommelier would at a Joselito tasting, attentive, knowledgeable, and never transactional. It operates in two distinct modes depending on who is interacting.
Consumer concierge
Guides customers through product discovery with personalised recommendations based on occasion, taste preferences, and budget. Provides slicing tutorials, storage guidance, wine and cheese pairings, and gifting suggestions, in Spanish, English, and Chinese. Designed to feel like a knowledgeable friend, not a chatbot.
Professional operations portal
A dedicated mode for chefs, gourmet retailers, and distributors, accessible via a secure login. One-click downloads for certifications, traceability reports, and branded marketing materials. Automated staff training content and kitchen posters. Eliminates the manual email chains slowing down Joselito's B2B relationships.
WhatsApp & WeChat integration
Meets international buyers on the channels they already use. WeChat integration was built into the MVP spec, not as a future add-on, because Joselito's fastest-growing market is Asia, where luxury food gifting is a serious cultural practice and WeChat is the default commerce channel.
Internal analytics dashboard
Tracks most asked questions, revenue impact by conversation flow, bot-assisted conversion uplift, and AOV trends, turning service conversations into continuous product and marketing intelligence.
Prototype
The prototype covered five screens and demonstrated the full user journey: the website hero, the branded concierge chat interface, the internal analytics dashboard, the For Professionals landing page, and the Master Intelligence portal showing all three B2B modules side by side.
My contribution
I owned this pillar from brief to prototype
This was a group capstone project at IE Business School, submitted December 2025. Within the team, I was solely responsible for the Digital Waiter pillar, identifying the strategic gap, designing the full product architecture across both consumer and professional touchpoints, and building the Figma prototype used in the final presentation.
Product strategy
Defined the strategic case for an AI-powered concierge: mapped the digital customer journey friction points, benchmarked luxury AI implementations (Chanel, Burberry), and built the business case anchored to conversion uplift and AOV data from conversational commerce research.
UX architecture
Designed the full dual-mode system, one AI engine operating as a consumer concierge on the website, and as a professional operations portal for chefs, distributors, and retailers, with distinct interaction patterns for each user type.
Figma prototype
Built five screens: the website hero, the branded chat interface, the internal analytics dashboard, the For Professionals landing page, and the Master Intelligence portal with three B2B modules. Presented live to the CEO of Joselito.
Key product decisions
Choices that shaped the design
Tone before features
I wrote the AI persona brief before designing a single screen, warm, refined, knowledgeable, never salesy. For a luxury brand, how the AI speaks matters more than what it can do. Tone guidelines and example response scripts were defined first and used to evaluate every UX decision that followed.
One engine, two distinct UX modes
Rather than building two separate tools, I designed a single underlying system that dynamically adjusts its context based on user type. Consumers get a concierge experience. Professionals get an operations portal. Brand voice stays consistent, infrastructure stays lean.
China-first multilingual architecture
WeChat integration and Chinese-language support were built into the core MVP spec, not retrofitted later, because the Chinese luxury consumer needed a channel they already trust, not a website with a language toggle.
Phased MVP with clear decision gates
90-day Spanish pilot → 6-month European expansion → 12-month global rollout. Each stage had measurable go/no-go checkpoints rather than a single big-bang launch, de-risking the investment for Joselito.
What made it different
Why this wasn't just a chatbot
Brand-first interaction design
The persona, tone, and even the refusals, what the AI does not say, were designed to feel like a Joselito staff member. Elegance and restraint over information density.
Built for global from day one
Trilingual support and WeChat integration were in the MVP, not the roadmap. International buyers were treated as primary users, not an afterthought.
B2B as a first-class product
Designing the professional portal as a real product, not an admin panel, made the immediate ROI case far stronger and addressed the most underserved users in Joselito's ecosystem.
Service that generates data
Every conversation produces signal, what customers ask, where they hesitate, which pairings convert. The AI turns customer service into a continuous product research channel.
The insight that shaped everything: Joselito's value is not just the product, it is the knowledge and ceremony around it. Every cut has a ritual. Every pairing has a reason. The problem was that none of that translated online. The Digital Waiter was the solution to making that expertise available everywhere Joselito's customers already were.
Outcomes
What the strategy was built to deliver
Projected 8–12% online conversion uplift in AI-assisted sessions, validated against luxury conversational commerce benchmarks from McKinsey, Forrester, and Gartner (2023–24).
Projected 15–20% increase in average order value on AI-influenced purchases, driven by pairing recommendations and guided upsells at natural decision moments.
Estimated €200,000–€400,000 in incremental annual revenue from the AI pillar alone, assuming 50,000 annual online visits with 25% AI adoption and a 20% basket uplift.
Projected 30–40% reduction in repetitive B2B support queries, freeing Joselito's team from manual communications and enabling consistent partner service at international scale.
Prototype was presented live to the CEO of Joselito in the final capstone session, the concept and interaction design received a positive response from the client stakeholder.
What I learned
Four things this project sharpened
AI product design is brand design
Building an AI experience for a luxury brand is not a technical challenge, it is an identity challenge. The persona, tone, and what the AI chooses not to say matter as much as the features it has.
MVP scoping is strategy
Choosing what to exclude from V1 is as important as what to include. A scoped MVP with clear KPIs and decision gates is far more persuasive to stakeholders than a fully featured concept with no test path.
B2B users deserve great UX too
Professional partners are often the most underserved users in a brand's digital ecosystem. Treating the B2B portal as a first-class product unlocked the strongest ROI argument in the entire project.
Heritage is content, not a constraint
Joselito's six-generation story was the raw material for every great AI interaction, acorn harvests, curing time, the exact angle of the knife. The brand's past was the product's best feature.