A UX exploration by Tom Oliva — showing how conversational AI could extend Félix's WhatsApp-first experience to the web, helping users understand the service before their first transfer.
This assistant knows Félix's service end to end — how transfers work, which countries are supported, pickup locations, payment methods, and security. It responds in English or Spanish, based on what you write.
The bot never pretends to process real transfers. It always redirects to WhatsApp for actual transactions — transparent about its limitations.
Detects and responds in the user's language automatically — key for a product serving Spanish-speaking migrants from a US-based interface.
Reduces the cognitive load of a blank input — especially important for users who may be unfamiliar with chatbots.
Every response about security is direct and reassuring — matching the tone Félix uses across their brand ("confiable 100%").
Félix is already a conversational product — its core experience runs on WhatsApp. This concept explores how that same conversational model could extend to the web, helping users understand the service before their first interaction. The design challenges here are identical to the ones I solve in fintech UX every day.
Sending money internationally requires complete trust in the service. Every UX decision — from response tone to error handling — needs to reinforce reliability. I applied the same principles I used designing Shell Box's payment flows.
Félix's users may not be tech-savvy. A blank chat input with no guidance creates friction before the conversation starts. Suggested replies lower the entry barrier — same problem I solved in loyalty redemption flows.
A bot that tries to do everything fails at everything. This assistant is explicit about what it can and can't do — and redirects clearly. That transparency increases user confidence, not decreases it.
Félix serves Spanish-speaking users from an English-language infrastructure. The assistant auto-detects language per message — no toggle, no friction. The channel adapts to the user, not the reverse.
This isn't a replacement for Félix's WhatsApp flow — it's an extension. Users discover and understand the service here, then complete the transfer where Félix already lives.
This concept was designed and deployed as a working prototype in under a week. The speed came from applying existing UX knowledge to a new channel — not from learning everything from scratch.
Félix Pago is a remittance service that operates 100% through WhatsApp — Latino immigrants in the US send money to Latin America without downloading any app. The product is conversational at its core: WhatsApp is the interface, not just the channel.
The design opportunity: potential users need to understand the service before initiating their first WhatsApp transfer. The existing help center was article-based — passive, search-dependent, not optimized for someone new to the product.
Build a web-based conversational assistant that mirrors Félix's WhatsApp interaction model, reduces pre-transfer friction, and demonstrates how conversational UX principles apply to fintech and remittance contexts.
In a traditional UI, users navigate menus and forms. In Félix's model, the interface is dialogue. The same UX principles — progressive disclosure, error prevention, trust signals — apply to conversation design. Instead of screen states, you design turns. Instead of button labels, you write utterances. The translation is direct.
12 core intents mapped from the actual Félix Pago help center and user journey, covering the full transfer lifecycle plus support scenarios.
| Intent ID | User utterances (examples) | Entities | Priority | Fallback path |
|---|---|---|---|---|
| HOW_TO_SEND | "How do I send money?" / "¿Cómo envío dinero?" | — | High | General_Fallback |
| FEE_INQUIRY | "How much does it cost?" / "¿Cuánto cobran?" | country, amount | High | Fee_Clarification |
| TRANSFER_TIME | "When does the money arrive?" / "¿Cuánto tarda?" | delivery_method | High | General_Fallback |
| COUNTRY_CHECK | "Can I send to Peru?" / "¿Envían a Colombia?" | country | High | Country_NotAvailable |
| PAYMENT_METHOD | "How do I pay?" / "¿Aceptan efectivo?" | payment_type | High | General_Fallback |
| SECURITY_TRUST | "Is Félix safe?" / "¿Es seguro?" | — | Medium | General_Fallback |
| SEND_LIMIT | "How much can I send?" / "¿Cuál es el límite?" | verification_level | Medium | General_Fallback |
| CANCEL_REFUND | "Can I cancel?" / "¿Puedo cancelar un envío?" | — | Medium | Escalate_Agent |
| PICKUP_LOCATIONS | "Where can they pick up cash?" / "¿Dónde retiran?" | country, city | Medium | General_Fallback |
| TRACK_TRANSFER | "Where's my transfer?" / "¿Ya llegó?" | — | Redirect | Escalate_WhatsApp |
| CONTACT_AGENT | "Talk to a person" / "Hablar con agente" | — | Redirect | Escalate_Agent |
| OUT_OF_SCOPE | "Can I send a top-up?" / "¿Hacen recargas?" | — | Fallback | OOS_Clarification |
TRACK_TRANSFER and CONTACT_AGENT are redirect intents, not informational ones. The bot cannot access real account data — the correct UX decision is transparency about that limitation and a direct route to WhatsApp. Pretending to have access would erode trust.
Every fallback ends with a concrete next action — an alternative within scope, or a direct path to a human agent. The bot never leaves the user in a conversational dead end. This directly addresses the core finding from the LATAM diagnostic: "no escalation at the right time" is the primary driver of conversational frustration.
Félix Pago serves Latino migrants in the US — people sending money home, often with anxiety about whether the transfer will arrive safely. The voice needs to be warm and trustworthy, not patronizing or corporate.
Target KPIs defined for post-launch measurement. Since this is a working prototype, these represent the measurement methodology I'd implement and present to stakeholders.
CANCEL_REFUND has the highest error rate — 6% of real cancellation requests route to OUT_OF_SCOPE. This is a high-stakes misclassification. Fix: expand cancellation utterance examples and add a disambiguation prompt when confidence is below 0.7 threshold.
Before deploying the AI-powered prototype, I'd validate conversation design using Wizard of Oz (WOz): a researcher manually responds to user inputs in real time using a prepared response library. The user believes they're interacting with an automated system. This surfaces dialogue issues before expensive model training.
| Block type | Usage | Key settings |
|---|---|---|
| Speak | All bot responses | Language: ES/EN via variable. Max 3 sentences per block. |
| Intent | Capture user input after each speak | 12 intents. Confidence threshold: 0.7. Below → Clarification block. |
| Condition | Route by intent, check entity presence | If country=null AND intent=FEE_INQUIRY → Fee_Clarification flow |
| Variables | user_language, detected_country, session_intent, fallback_count | fallback_count increments on unrecognized input. At 3 → force escalation. |
| API Block | Call Claude API for dynamic responses | System prompt with Félix knowledge base. Max tokens: 300. |
| Buttons | Suggested replies (3 max) | User-perspective options only. Never "Tell me more." |
| No Match | Input doesn't match any intent ≥0.7 | 1st miss: clarify. 2nd miss: offer categories. 3rd miss: escalate. |
| No Reply | User idle >30 seconds | "Still there? Take your time — or start fresh when you're ready." |
This case demonstrates something I strongly believe: conversational UX is not a separate discipline from product design. The core skills are identical — understanding user goals, defining clear states and flows, handling errors gracefully, building trust through transparency.
What's different is the medium. Instead of screen states, you design turns. Instead of button labels, you write utterances. The experience designing fintech flows at Shell Box and Ewally mapped directly onto this conversational context.
I designed this assistant end-to-end — from intent architecture to voice principles to a working deployed prototype — in under a week, using existing product design skills applied to a new channel. That's not a gap in experience. That's evidence of transferable craft.