In this interview, e27 speaks with Bell Beh, CEO and Co-Founder of Buzz, a travel payment platform focused on building digital infrastructure for underserved cross border travellers, starting with APAC-MENA and Muslim friendly travel. As AI capabilities expand across industries, Buzz is applying vertical AI models to address the unique needs of halal travel, pilgrimage logistics, and cross-border payment experiences across APAC and MENA.
This conversation forms part of e27’s broader AI Pulse coverage, which examines how organisations across the region are building, deploying, and governing AI in real-world settings.
Building AI companions for Muslim travel
e27: Briefly describe what your organization does, and where AI plays a meaningful role in your work or offering.
Bell: Buzz is a travel payment platform, specialising in building the Muslim travel infrastructure. At the core of our offering is BAE (Buzz AI Experiences), wee have built three AI travel companions, AI Tour Guide (BAE for Global Travel), first use case was launched with the Singapore Land Authority in 2024, Omani BAE, and Hajj & Umrah BAE — each designed to support different Asian travel needs from leisure to pilgrimage.
AI is crucial in delivering halal-aware recommendations, prayer-time–sensitive itineraries, multilingual assistance, and culturally relevant support across APAC,MENA and other unfamiliar territories. It also powers smart booking flows and secure cross-border payment optimization to support Muslim travellers globally.
Real-time guidance and booking integration
e27: What is one concrete way AI is currently creating value within your organisation or for your users or customers?
Bell: Our AI Tour Guide (Global BAE) creates value by giving travellers contextual tips and recommendations in real time, then guiding them straight into booking and payment flows globally. For faith-based journeys, Hajj & Umrah BAE provides halal-aware guidance and dynamically adjusts plans around prayer times, nearby halal dining, visa requirements, and local customs. This reduces friction and uncertainty during travel, improves user confidence, increasing booking conversion, and driving higher payment transaction volume on our platform.
Also read: AI Pulse Exclusive: How CoBALT is designing AI that teams can actually trust
Prioritising domain-specific AI models
e27: What was a key decision or trade-off you had to make when adopting, building, or scaling AI?
Bell: A key decision we made was choosing to build domain-specific AI models for Muslim travel instead of relying purely on large generic LLMs. This meant slower initial scaling and higher training effort, but it ensured cultural accuracy, halal compliance, and pilgrimage-specific intelligence that generic models often miss.
We prioritised depth and trust over speed, because in faith-based travel and cross-border payments, precision and contextual relevance matter more than volume.
Conversion gains and regulatory complexity
e27: Looking back, what has worked better than expected, and what proved more challenging than anticipated?
Bell: What worked better than expected was how quickly AI-driven personalization translated into measurable conversion and transaction uplift, especially when embedded directly into booking and payment flows. Our direct partnership with the Indonesian Hajj authority (BPKH) further validated that purpose-built AI for pilgrimage travel is not just a consumer feature, but institutional-grade infrastructure.
More challenging than anticipated was aligning AI deployment with regulatory, compliance, and cross-border payment requirements across multiple jurisdictions. In faith-based and financial contexts, accuracy, trust, and governance standards are significantly higher than typical travel use cases. Also, both travel and payment are heavily regulated, so applying licenses on both sides to reach the Agentic AI use case was challenging, especially when we are building the infrastructure layer.
AI must align with regulation and trust
e27: What is one lesson about applying AI in real-world settings that leaders or founders often underestimate?
Bell: One lesson founders often underestimate is that AI must align with regulation and operational reality, not just product vision. In sectors like travel and payments for us at Buzz, AI cannot bypass licensing, compliance, or trust requirements — it must integrate with them.
Agentic AI may look seamless in demos, but in the real world, automation is only as strong as the regulatory rails and human oversight behind it.
Also read: AI Pulse Exclusive: How Explico is building AI teachers can actually rely on
Building AI through vertical use cases
e27: Based on your experience, what is one practical recommendation you would give to organisations that are just starting to explore or scale AI?
Bell: Start with a clear vertical use case, not a generic AI deployment. AI delivers real value when it is deeply embedded into a specific workflow — whether that’s booking, payments, or customer guidance — rather than layered on as a chatbot.
Treat it as a 3-year progression, and don’t expect AI to be a magic pill: Year 1 focus on data readiness, safety, and assisted experiences; Year 2 integrate AI into core operations with human-in-the-loop and compliance-by-design; Year 3 scale toward more agentic automation once regulatory rails, partnerships, and governance are proven.
In a progressive manner, solve one high-friction problem end-to-end, ensure it aligns with compliance and operational realities, and scale from there.
From tour guide to agentic travel assistant
e27: Over the next 12 months, how do you expect your organisation’s use of AI, or the role of AI in your industry, to evolve?
Bell: Over the next 12 months, we expect AI to shift from being primarily an AI Tour Guide to becoming a Agentic AI Travel Agent that can execute—within regulated boundaries—more of the Plan – Decide – Act – Execute. It‘s no longer just suggestiona/ assisting with travel itineraries. It’s execution layer for our Vertical AI trained since 2024 with various governments.
For Buzz, this means deeper integration of BAE into partner inventory and cross-border payment rails across APAC–MENA, with stronger compliance, auditability, and human-in-the-loop controls.
Across the industry, AI will move from chat-based experiences to workflow-native automation, but adoption will be paced by regulation, trust, and licensing readiness—especially in travel fintech and Muslim travel infrastructure.
Also read: AI Pulse Exclusive: How Spacely AI is bringing generative AI into spatial design workflows
Final thoughts on vertical AI
e27: Anything else you want to share with the audience?
Bell: 2026 is the year of vertical AI, choose one niche and execute well.
The rise of vertical AI in specialised sectors
This conversation highlights how vertical AI is emerging across specialised sectors where cultural context, regulatory compliance, and domain expertise are critical. In areas like Muslim travel and cross-border payments, AI systems increasingly function as infrastructure embedded into booking, guidance, and transaction workflows. As technology evolves, organisations may find that the most durable advantage comes from combining deep domain knowledge with carefully governed AI systems.
For more interviews, analysis, and real-world perspectives on how organisations across the region are applying AI in practice, click here.
—
Enjoyed this read? Don’t miss out on the next insight. Join our WhatsApp channel for real-time drops.
The e27 team produced this article
We can share your story at e27 too! Engage the Southeast Asian tech ecosystem by bringing your story to the world. You can reach out to us here to get started.
Featured Image Credit: Buzz
The post AI Pulse Exclusive: How Buzz is building AI-powered infrastructure for Muslim travel appeared first on e27.