KW

THAILAND TOURISM · PRODUCT HACKATHON

Thai Unseen

A bilingual, multi-sided tourism business concept connecting traveler demand with hidden gems, local guides, community businesses, travel buddies, and transparent transport.

AI Travel Product Hackathon
Thai Unseen
4 SIDEStraveler and provider ecosystem
TH / ENbilingual experience
LIVEinteractive prototype

The problem behind the interface.

The challenge was broader than designing another destination directory. Thai Unseen explores a multi-sided platform where travelers receive trustworthy discovery and planning, while local guides, drivers, and community businesses gain visibility and access to demand. I translated that business hypothesis into a responsive, bilingual prototype with connected discovery, matching, provider, buddy, and transport experiences.

What I owned,
not only what the team shipped.

Responsibilities are written as decisions and deliverables so the individual contribution stays visible.

01

Mapped the value proposition across travelers, local guides, community businesses, and transport providers.

02

Translated the business concept into a connected traveler journey from inspiration through planning and local connection.

03

Built the responsive Next.js / TypeScript interface and reusable product components.

04

Implemented the route-matching interface and explainable scoring prototype around traveler constraints.

05

Connected discovery, provider supply, transparent pricing, and bilingual content into one demonstrable product.

From raw signal
to useful response.

A simplified architecture showing how information changes as it moves through the system.

01

Attract

Traveler intent and hidden demand

02

Match

Places, guides, businesses, and rides

03

Connect

Trusted providers and fair pricing

04

Circulate

More value reaches local communities

FrameworkNext.js / TypeScript
InterfaceTailwind CSS / Motion
BusinessMulti-sided marketplace concept
LocalizationCustom TH / EN dictionary

Designed around
the real constraint.

Features matter here because each one answers a workflow, data, or operational constraint.

01

A multi-sided tourism business ecosystem

Framed distinct value for travelers, local guides, community businesses, and drivers: trusted discovery for demand, and visibility plus booking opportunity for local supply.

A multi-sided tourism business ecosystem
02

Route matching around the traveler

Built a route-search prototype that scores places using the search intent, starting area, interests, budget, distance, and route fit.

Route matching around the traveler
03

Hidden gems as a discovery experience

Organized lesser-known destinations into an approachable visual discovery flow across nature, culture, beaches, history, and food.

Hidden gems as a discovery experience
04

Local connection, not only destination content

Extended the journey beyond places with verified local guides, travel-buddy matching, trip planning, and transparent transport concepts.

Local connection, not only destination content
05

Bilingual, reusable product foundation

Structured reusable React components and a custom TH / EN dictionary so the prototype could present one coherent experience to domestic and international travelers.

What can be claimed.
What still needs proof.

Useful engineering separates observed evidence from interpretation and future validation.

LIVEPrototype

The core product journeys are deployed and can be explored interactively.

4Market sides

Tourists, guides, businesses, and drivers have distinct entry points and value propositions.

TH / ENLocalization

The interface supports Thai and English through a reusable dictionary structure.

Known limitations

  • Route recommendations currently use curated data and explainable rule-based scoring rather than a learned recommendation model.
  • Guide verification, booking, messaging, authentication, and payments are prototype flows rather than production integrations.
  • Revenue model, provider acquisition cost, take rate, and marketplace liquidity remain unvalidated business hypotheses.
  • Recommendation relevance and usability have not yet been validated with a representative traveler study.

Next engineering steps

  • Evaluate a hybrid recommender combining traveler constraints, semantic place retrieval, and transparent ranking signals.
  • Validate willingness to pay, provider onboarding, transaction economics, and an appropriate commission or partner-revenue model.
  • Persist traveler preferences, providers, and trip state in a production data model with secure access controls.
  • Run task-based usability and recommendation-quality evaluation with Thai and international travelers.
NEXT CASE STUDY

Piston Defect Intelligence System