Wiraphat Yodsri

Architected a non-disruptive debugging service (FLAPI Debug Tool) that isolates debug processes from the production environment; this was achieved by enabling endpoints exclusively on a mirror environment and utilizing an asynchronous polling mechanism to retrieve cached data post-transaction, ensuring zero impact on live services.

Designed and implemented a comprehensive logging pipeline to capture and store flight fare rules and free baggage allowance information from over 100 suppliers in Hadoop, making the data accessible through an internal tool for the Flights Department

Engineered a real-time data mapping solution to display large-volume log messages (over 4MB for round trips, 0.4MB for one-way) on the FLAPI Debug Tool, mirroring the data logged to Hadoop for immediate analysis

Orchestrated the migration of the Flights Support Tool deployment from Kubernetes manifests to Agoda's private cloud platform (Buckbeak), establishing a complete CI/CD pipeline with automated testing, logging, and monitoring.

Streamlined the development lifecycle by onboarding devstack for the FLAPI Debug Tool, creating manifests that enabled developers to deploy to a development environment with a single click in the GitLab CI pipeline.

Integrated a Meta Search Engine (MSE) feature into the internal debugger service, empowering Product and BI teams to directly access and debug live data from all suppliers, significantly streamlining their workflow.

Authored comprehensive unit and component tests for the FLAPI Debug Tool's frontend, achieving 96% branch coverage and enhancing application stability.

Resolved a critical logging manager bug that caused data loss by implementing a data-merging technique to combine existing and new cache information, ensuring the accurate display of flight data from all suppliers.

Built an API service integrating AI to summarize patient's medical histories, reducing physician's manual review time by 30%.

Created comprehensive unit tests for backend services, achieving 95% code coverage and ensuring robustness and system reliability.

Led the full-stack development of AI-powered tools, overseeing the entire software development lifecycle from initial concept to final deployment, ensuring timely delivery and system scalability.

Developed a specialized API service that enabled AI-driven style transfer, converting static interior design images into AI-enhanced visuals with new design elements, improving user experience and creativity.

Designed and optimized user interfaces for seamless interaction with AI tools, allowing users to visualize real-time interior design changes such as changing image to video, customizing room perspective and room styling, driving engagement and enhancing usability.

Designed and developed middleware for user authorization, integrating Microsoft Azure AD token verification and implementing Role - Based Access Control(RBAC) to manage permissions within the Content Management System(CMS).

Led the frontend development of the CMS platform, utilizing global state management to enhance performance, reducing unnecessary re-renders and improving user experience.

Integrated Axios with Tanstack Query to efficiently manage API calls, implementing error handling, data caching, and reducing redundant requests, resulting in faster load times and improved data reliability.

Developed the customer service system for handling post-sales product repairs for Bewell, utilizing Golang as the backend server.

Designed the database for the customer service system using MS SQL, implemented indexing to enhance query performance, and defined relationships according to database design principles.

Designed the customer service system with a strong focus on architecture, using design patterns like Singleton and Factory, implementing clean architecture, and adhering to SOLID principles and DDD.

Developed the delivery sales system for Bewell, utilizing Golang as the backend to ensure efficient and reliable operations.

Implemented the backend system using Golang as the server and employed gRPC as the API architecture to facilitate communication between servers, effectively reducing the latency of API calls to support over 1,000 users.

Authored proto files according to the API requirements to integrate and establish the gRPC server, allowing clients to access and interact with the services efficiently.

Designed the code architecture by implementing hexagonal architecture principles, structured the codebase in a domain-based manner, and utilized Golang as the backend server to ensure a modular and maintainable system.