Yuva R

• Led a dynamic Machine Learning implementation team responsible for seamlessly integrating ML models into various data processing pipelines. Spearheaded the design and execution of integration process, ensuring smooth incorporation of trained ML models into existing production systems.
• Led technical backlog grooming for ML model integration through collaboration with cross functional teams such as Data Science, Product Management and Engineering.
• Conducted thorough benchmarking of machine learning models, optimizing their deployment on ECS to achieve optimal throughput-to-cost ratio. Proficiently leveraged g4dn instance types for GPU-based inferences, enhancing the overall efficiency of ML deployment.
• Designed a robust ML Quality Assurance (QA) system using AWS Batch and Python Django, ensuring reliability and accuracy of ML models in real-world scenarios.
• Architected and constructed end-to-end data pipelines and inference pipelines on AWS cloud, employing NLP-based ML models and Apache Spark to enhance data processing capabilities.
• Designed and built data analysis, data validation and inference pipelines using SageMaker Studio notebooks and EMR clusters.
• Implemented a streamlined version control system by setting up DVC (Data Version Control) with AWS S3 as the backend. Effectively configured it to version model artifacts in conjunction with the source code, promoting a systematic approach to model development and maintenance.

• Led a dynamic Machine Learning implementation team responsible for seamlessly integrating ML models into various data processing pipelines. Spearheaded the design and execution of integration process, ensuring smooth incorporation of trained ML models into existing production systems.
• Led technical backlog grooming for ML model integration through collaboration with cross functional teams such as Data Science, Product Management and Engineering.
• Conducted thorough benchmarking of machine learning models, optimizing their deployment on ECS to achieve optimal throughput-to-cost ratio. Proficiently leveraged g4dn instance types for GPU-based inferences, enhancing the overall efficiency of ML deployment.
• Designed a robust ML Quality Assurance (QA) system using AWS Batch and Python Django, ensuring reliability and accuracy of ML models in real-world scenarios.
• Architected and constructed end-to-end data pipelines and inference pipelines on AWS cloud, employing NLP-based ML models and Apache Spark to enhance data processing capabilities.
• Designed and built data analysis, data validation and inference pipelines using AWS SageMaker Studio notebooks and EMR clusters.
• Implemented a streamlined version control system by setting up DVC (Data Version Control) with AWS S3 as the backend. Effectively configured it to version model artifacts in conjunction with the source code, promoting a systematic approach to model development and maintenance.

• Revitalized the performance and scalability of the existing system, contributing significantly to the seamless onboarding of new customers. Instrumental in enhancing system capabilities to accommodate increased demand and ensure optimal user experience.
• Successfully designed and implemented a robust data serving API, fostering efficient communication and data retrieval within the system. This API implementation played a key role in elevating the overall functionality and user satisfaction of the platform.
• Demonstrated a hands-on approach in the deployment and ongoing maintenance of the system in a production environment. This involvement ensured the reliability and continuous operation of the system, meeting the evolving needs of users and stakeholders.
• Expertly utilized a diverse technology stack, including AWS, Python, Django, Java, Spring Boot, RESTful Web services, Docker, Couchbase NoSQL database, Git, and Maven. This comprehensive skill set facilitated a well-rounded and versatile approach to system development and maintenance.

Experience:

• Acquired and applied new skills, including Docker, Couchbase NoSQL Database, and Microservices, through hands-on learning during the course of software development.
• Championed the orchestration of microservices using the Docker platform, streamlining the deployment and management of modular components within the system. This initiative significantly contributed to the system's scalability and maintainability.
• Translated system specification documents into practical solutions by designing and implementing SOAP web services. This involved a meticulous process of understanding requirements and architecting solutions to meet the specified system functionalities.

Skills:

• Spring Framework, Spring Boot, REST Web services, Docker, Couchbase NoSQL Database, Couchbase SyncGateway, Git, Maven.
• Java, J2EE, JBoss EAP, SOAP Web services.

Projects:

• RDA - RailDOCS Anywhere is a mobile solution for managing, testing and inspecting wayside assets of US railroad customers.
• SAMP - Stand Alone Movement Planner is an advanced railroad network optimisation software system that provides real-time traffic planning and network optimization. It detects meet/pass conflicts and optimizes routes. With Movement Planner, customers can move freight faster, at low cost, using existing rail networks.

Worked on customer service and support tickets for infrastructure and applications' maintenance as part of the Devx project.

Experience :

• Contributed to development and delivery of 4 projects
• Learnt new skills on the job and effectively used them during software development.
• Supported UAT activities at customer location (Hong Kong)

Skills:

• Java, J2EE, Spring Framework, iBatis ORM, Oracle 10g, PL/SQL, Documentum.

Projects:

• BVPTF - Product Technical Folder is an application based on Content Management. It deals with Versioning, workflow management, lifecycle management for the documents required during the product lifecycle.
• BVPRICING - A financial guidance tool for the Bureau Veritas
• BVcpnetV2 - A web application for the clients of Bureau Veritas to place orders and for the vendors to verify those orders. It also generates multitude of business intelligence reports.
• MFUSOEOL - A web application to manage, synchronise and monitor ECUs installed on trucks for the MFTBC (Mitsubishi FUSO truck and Bus Corporation).