Certificates
Working as a Solutions Architect to design and implement cloud solutions.
Achieved 90% accuracy in part demand forecasting using Amazon Forecast. Developed a multi-agent system with LangGraph and built a video translation platform integrating open-source LLMs with Gradio.
Assisted over 80 students in obtaining AWS certifications through YouTube courses. Conducted workshops on CloudFront and Lambda@Edge, achieving 4.9 satisfaction rating. Engineered microservice architecture for email workflow automation, reducing processing time by 40%.
During my internship, I utilized AWS Glue and PySpark to build a data pipeline and perform a series of ETL processes, reducing data processing time by 50%. I also leveraged AWS Personalize to enhance marketing campaigns and recommendation systems, boosting targeted promotion accuracy by 20%. Additionally, I constructed a recommendation chatbot with LangChain and Amazon Bedrock LLM, increasing customer engagement and satisfaction by 30%.
In this role, I identified over 100 critical bugs through manual testing of BVT, SVT, and SFT for FaceMe Product. I conducted comprehensive API testing for FaceMe Security, enhancing endpoint functionality by 35%. Furthermore, I produced detailed test reports, expediting issue resolution and improving overall product quality.
Hosted SageMaker workshop on ML fundamentals and LLM fine-tuning with QLoRA, achieving 5.0/5.0 attendee satisfaction. Led hands-on sessions covering model deployment and optimization techniques, empowering participants with practical ML skills.
Provided technical support at the Card Crash booth, assisting attendees in challenging and exploring AWS's classic architectures. Facilitated interactive learning sessions and guided participants through complex cloud scenarios, resulting in enhanced understanding of AWS services.
Deployed resources using Serverless Framework for scalable deployment. Established automated data pipeline for PDF embedding into knowledge base, implemented LLM logic using LangChain, and developed an interactive Gradio frontend interface. The solution demonstrated innovative use of AI technologies in practical applications.
During the Microsoft competition, we harnessed Azure services to develop a demo showcasing key functions for the Assistive Shopping App for the Visually Impaired. This app offered product detection, information extraction from labels, and text-to-speech conversion, enhancing the shopping experience for the visually impaired.
I gained valuable experience in the AWS Serverless Data Analysis Workshop, covering S3-based serverless analysis with Athena, visualization via QuickSight, serverless ETL and data discovery with Glue, and data analysis in S3 using Redshift Spectrum.
Part of learning samples and personal projects. View more.
