Almosafer (Seera Group) · Travel Tech · AI · 2023

MENA's First
AI
Travel
Experience

Conceived and shipped the first LLM-powered travel booking experience in the Middle East, natural language search embedded in the Almosafer mobile app and scaled to WhatsApp, driving a 41% lift in search-to-booking conversion.

Role

Director of UX & Product Design

Timeline

2020 – 2023

Team Scale

12 → 25+ designers

Platform

iOS · Android · WhatsApp

+41%
Search-to-booking conversion
3.2x
More destinations explored
28K+
WhatsApp AI bookings in 3 months
Conversational Booking, WhatsApp + Backend Integration
CHANNELWhatsApp Business API
E2E LATENCY2.3s avg
MARKETSSaudi Arabia, UAE, Egypt
Almosafer AI
Always online
Family trip to Maldives, 4 nights, budget around AED 4000 per person
2:14 PM
Found 3 great options for your family! ✓
Centara Grand Maldives
4 nights · Flights included
AED 4,200
✓ In budget
Direct flight
Confidence: 96% match
Kandooma Resort
4 nights · Ferry transfer
AED 3,750
Confidence: 88% match
Tap any option to reserve →
2:16 PM
✓ Booking confirmed
Ref: ALM-482156
Confirmation email sent
2:16 PM
BACKEND PROCESSING
1. Webhook Receive (t+0ms)
WhatsApp message validated, user context loaded
2. NLU Parsing (t+240ms)
Destination, dates, budget, preferences extracted
3. Inventory Query (t+850ms)
Live flight/hotel APIs queried in parallel
Top 12 options ranked by fit score
4. Response Generation (t+1280ms)
LLM writes natural response + formats results
✓ Send to WhatsApp (t+1840ms)
Total: 1.8s end-to-end
Confidence Scoring
• Intent match: 96%
• Budget fit: 94%
• Availability: 98%
• Overall: 96% ✓
Channel Performance Comparison
WhatsApp
• 2.3s latency
• 87% completion rate
• 41% higher conversion
In-app Chat
• 1.6s latency
• 78% completion rate
• Baseline comparison
Traditional Search
• Form-based
• 54% completion rate
• 3.2x less discovery
Error Handling & Fallbacks
Ambiguous intent: AI asks clarifying questions in conversation
No availability: Shows alternatives + suggests date shifts
API timeout: Retries up to 3x, degrades to keyword search
Payment failure: Booking link persists for manual retry
2.3s
E2E Latency
87%
Completion Rate
+41%
Conversion Lift
3.2x
Destination Discovery
96%
Avg Confidence

Side project
to
production.

I initiated and led this end-to-end, from concept through production launch and WhatsApp scale-up. What started as a side project became Almosafer's most significant innovation bet.

01

Design sprint facilitation, Ran a focused sprint with senior members of my 25-person design org. Led context, framing, and solution direction. Within 5 days: validated concept, interaction model, and visual prototype.

02

Conversational UX architecture, Designed the complete natural language search experience: conversation initiation, AI clarification of ambiguous requests, results presentation within conversational flow, refinement ("show me something cheaper"), and booking handoff.

03

LLM-to-inventory integration design, Worked with engineering to connect the LLM to live flight and hotel APIs so every response contained real availability, real pricing, and real booking capability.

04

WhatsApp scale-up, Recognising WhatsApp handled 70%+ of customer interactions, designed the extension of AI search to WhatsApp. Customers could have the same natural language conversation, receive options with images and pricing, and trigger bookings via deep link to the app.

05

150-participant cross-market validation, Tested across Saudi Arabia and UAE. 87% of participants preferred natural language over form-based search for initial trip exploration. Customers explored 3.2x more destinations with AI search.

"We proved that LLMs could drive real commercial outcomes in travel, not just engagement metrics. Customers discovered places they'd never have considered if forced to type a destination into a dropdown."

Regional
first.
Commercial
impact.

Almosafer became the first travel platform in MENA to ship a production LLM-powered booking experience, establishing regional innovation leadership and proving the commercial viability of AI search.

Conversion Lift

41% increase in search-to-booking conversion for AI-assisted sessions vs. traditional search. Across all channels, omnichannel conversion improved by +110% over the same period.

Discovery

Customers explored 3.2x more destinations per session when using natural language search.

WhatsApp Scale

28,000+ AI-assisted bookings processed via WhatsApp within the first 3 months.

Time Savings

62% reduction in time-to-booking, from 12 minutes average to 4.5 minutes for AI-assisted sessions.

Next Case Study

Ask Careem:
AI-Powered Dineout Discovery

Want in?

Score 100% on the Design Eye test to earn access. Or enter the password if you have it.

4 rounds. No partial credit.