AI-Driven Recommendation System

AI-Driven Recommendation System is an end-to-end product that helps users quickly find suitable migration agents through AI-powered matching. The platform supports agent discovery, guided questionnaire input, and recommendation generation with ranked results, enabling users to reach out to agents with minimal friction.

I served as the solution architect and primary implementer for the entire system. I designed the overall architecture across frontend, backend, and the AI pipeline, including model research, data acquisition (crawler), data cleaning and preprocessing, feature engineering, model training and tuning, and finally integrating the best-performing model into production services.

On the engineering side, I delivered the full backend functionality and APIs, performed end-to-end integration with the UI, implemented self-testing and regression checks, optimized the frontend experience for performance and usability, and led the final project demonstration.

• Owned the full lifecycle: problem framing → architecture → data pipeline → model → backend → frontend → testing → presentation
• Conducted model research and benchmarking, selected final approach based on performance and deployment feasibility
• Built a complete data pipeline: web crawling, cleaning, normalization, preprocessing, and dataset preparation
• Implemented training and hyperparameter tuning workflow and integrated the final model into backend inference APIs
• Designed and delivered backend modules for agent profiles, search , filter, recommendation workflow, and contact flows
• Led API integration, end-to-end testing, and UI optimization, do product presentation to our client

Home page

Questions

Predict result

Project Management