Bajan Bridge
Helping visitors understand Bajan speech and use local advice through multilingual translation and visual transport guidance.
2nd place
BOLD Builders Tourism Innovation Hackathon · 2026
Team: Beyond the Trident
The problem
A visitor may struggle with Bajan expressions or understand a translated direction without recognising the vehicle or transport reference. Bajan Bridge connects language understanding with visual information that helps visitors interpret local advice.
My contribution
I helped conceptualise and implement Bajan Bridge with Beyond the Trident, working across both phases. I used AI-assisted development to help build the application and bring its translation and visual guidance features together.
Phase 1 · Multilingual translation
I worked on a translation workflow using Whisper, Qwen, and a dataset of Bajan sayings to support translation into Standard English, French, and Spanish. Whisper handles spoken input, while Qwen supports translation, with Bajan examples providing local language context.
Phase 2 · Visual guidance
I also helped implement recognition of key words in local responses so that relevant photos appear for the tourist. This adds visual context to the advice. For example, this includes helping a visitor recognise the type of transport being discussed.
Tools & methods
Python application development, AI-assisted implementation, Whisper speech recognition, Qwen through Ollama, a Bajan sayings dataset, keyword recognition, and image-based guidance.
Results
The team developed a demonstration application combining translation and visual assistance. Beyond the Trident earned second place at the BOLD Builders Tourism Innovation Hackathon on 25 September 2026.
What I learned
Building the two phases together highlighted that translating words is only part of helping a visitor: local context and recognisable visuals also matter. Working with the speech and translation tools showed how recognition errors, local expressions, and response speed affect the experience.
