Son Nguyen

Senior MLOps Engineer at ZALS, focusing on machine learning operations and infrastructure.

Worked as Senior MLOps Engineer at ZEALS.

• Lead the development of AI services like product generation, product recommendation.
• Help to plan to migrate services from VM based infraructure to container based infrastructure to save cost and improve scalability.

• Responsible for development and performance optimization for AI models in video-call
• Responsible for the enhancement and maintenance of lip-sync technology and related tools

• Designed and built a machine learning platform from scratch based on K8s.
• Optimized time to onboard new ML models from 2 weeks to 2 days, and 75% the cost of maintenance by applying MLOps practices.
• Optimized the computation cost for model training and data processing by 85% by leveraging spot instances.
• Achieved 99.9% availability SLI by building a monitoring and alerting system for our s

Using graph-based methods, develop a advanced friend ranking service used in wallet to wallet money transfer screen and home screen of Vietnam largest e-wallet application - MoMo, contributed to approximately 40,000 transactions per month
• Performed efficient preprocessing of voluminous raw data, applying cutting-edge graph-based algorithms to compute the friend ranking scores. The data scale enc

• Automated the integration of personalized recommendation engines into clients’ systems by developing an end-to-end recommendation platform using TypeScript and Azure’s services.
• Expanded client base by integrating our recommendation platform with partners like MoEngage and Haravan.

• Led the development of a similar item recommendation engine and employed Elasticsearch module for the search engine for the largest online store pharmacy chain in Vietnam.
• Implemented a novel search & suggestion box and seamlessly integrate the search engine into the existing website. This strategic enhancement led to a remarkable 250% surge in daily search queries and a substantial 45% increa

• Lead a team building online scoring and online examining system based on PHP, VueJS and OCR

• Undertook the responsibility of maintaining and advancing the ASP and DMP modules of a large-scale affiliate advertising system for both advertisers and partner websites. With backend was developed using technologies such as SpringBoot, SQL, and AWS services (DynamoDB, Redshift...) and frontend was developed using AngularJS.
• Proficiently maintained and developed message-based chat applications