Ritwik Dey

Amazon Q - Amazon Q, generative artificial intelligence (AI)-powered assistant for accelerating software development.

Leveraged AWS Batch for a distributed data ingestion pipeline, enabling real-time processing and analysis with scalability.
Ensured high-quality training data and improved generative model performance through advanced data filtering techniques like bias detection and PII identification.
Designed and deployed a Docker-based containerized architecture for model evaluation and benchmarking, enabling seamless scaling and fault tolerance.
Integrated CI/CD practices with automated testing, code reviews, and deployment pipelines for reliable software delivery.
Collaborated with cross-functional teams, gathering requirements and aligning objectives for successful project delivery.
Conducted regular performance monitoring and profiling, optimizing resource utilization and addressing bottlenecks.
Implemented robust security measures like data encryption, access controls, and auditing to protect sensitive training data and model artifacts.
Expanded the training dataset and improved model generalization through advanced data augmentation techniques like back-translation and text generation.
Developed custom evaluation metrics and benchmarking methodologies tailored to generative AI capabilities like coherence, fluency, and factual accuracy for comprehensive model quality assessment.

Digital International Pricing System

Digital International Pricing System is the global pricing system for all the digital ecommerce business and subscription like kindle and prime video.

Responsibility :

• Architected and implemented a unified pricing algorithm workflow, streamlining operations for subscription-based digital products across multiple businesses and marketplaces.
• Spearheaded the onboarding of 10 new businesses and seamlessly migrated approximately 26 existing businesses from legacy pricing systems to the new unified workflow, ensuring continuity and efficiency.
• Developed and deployed tailored pricing algorithms for à la carte digital businesses and products, catering to diverse product offerings and pricing models.
• Orchestrated the onboarding of new product lines and successfully launched pricing features for different global marketplaces, ensuring seamless integration with various pricing algorithms.
• Managed end-to-end operational activities related to pricing, encompassing pipeline maintenance, ticket support, dependency management, version upgrades, and proactive risk mitigation strategies.
• Implemented automated data archival and pricing anomaly detection systems, enhancing data governance and enabling proactive identification of pricing discrepancies.
• Collaborated cross-functionally with stakeholders, leveraging domain expertise to optimize pricing operations and drive continuous improvement initiatives.
• Established robust monitoring and reporting mechanisms to track pricing performance metrics, enabling data-driven decision-making and identifying optimization opportunities.
• Conducted regular audits and assessments to ensure pricing compliance with internal policies, regulatory requirements, and industry best practices.
• Developed comprehensive documentation and knowledge-sharing resources, facilitating seamless knowledge transfer and training for pricing operations teams.
• Implemented automated testing frameworks and continuous integration/continuous deployment (CI/CD) pipelines to ensure the reliability and scalability of pricing systems.
• Collaborated with pricing & vendor account managers and explored ,evaluated emerging pricing strategies, driving innovation and staying ahead of industry trends in pricing optimization.

LOT Order Management System : LOT Order Management product enables organization/business to manage whole life cycle of an order from creation to till merchandised.

Responsibility :

• Analyze the requirements, prioritize them based on business needs, and document them as functional and non-functional specifications. Break down the requirements into manageable user stories or features.
• Define project milestones and create a detailed timeline for deliverables. Allocate appropriate resources, such as team members and budget, for successful project execution.
• Organize and prioritize user stories or features based on the defined project milestones.
• Procure necessary infrastructure resources, such as servers, databases, and networking components, for the QA, development, and production environments.
• Create a comprehensive system design document outlining the overall architecture.
• Develop conceptual data models to represent business entities and their relationships. Based on these conceptual models and database requirements, create logical data models defining data entities, attributes, and relationships. Ensure a robust and well-structured data architecture for the application.
• Create wireframe diagrams for the user interface based on the user stories and requirements. Collaborate with business stakeholders to review and obtain approval for the wireframes.

Advance Manufacturing IoT Solutions : Advance Manufacturing is IoT solution for vehicle assembly line. It enables all assembly line- side applications and hardware(PLCs,robots) to capture and analyze the shop floor data and communicate with different IoT end point devices. It enables customer to create custom reports in data visualization tool, proactive monitoring and preventive maintenance of line side machines and robots.

Responsibility :

• Worked closely with the company to identify customer needs and demands of IoT solution and risk assessment of implementation for Global plants.
• Create Data Model and ER diagrams to store the plant floor machines data in database.
• Create Event triggers in IoT solution to communicate with line side machine, fetching machine data and update in database.
• Analysis of captured data, identify the data pattern, create visualization and report generation.
• Data analysis of captured data and review with business to understand the usage patterns and create reports for business review.
• Designed and developed different machine learning models on large data sets using SQL database to predict machine failure event. Successfully interpreted data to identify key metrics and draw conclusions and present to business.
• Proposed machine learning solutions to improve plant floor system efficiencies, leading to a 10% reduction in plant down time and efficient maintenance calendars.