Yura – Python, Django, AWS

AI-enabled insurance automation platform for large US commercial vehicle fleets, including ambulance services, transportation companies, and other fleet-based businesses. The platform covers application intake, pricing, policy generation, contracts, documentation, and analytics.

• Built an LLM-powered email-processing workflow with Claude that extracts and structures information from incoming fleet insurance applications and supports the application-to-contract process.
• Automated multi-step document and contract-processing workflows, reducing manual work across insurance operations.
• Owned end-to-end delivery across customer-facing and internal services, including React frontend, Python backend, integrations, infrastructure, and production support.
• Rebuilt service communication from synchronous HTTP calls to event-driven Apache Kafka messaging, removing backpressure and improving scalability.
• Built an asynchronous email-delivery service using Amazon SQS and SES for reliable background processing.
• Integrated Amazon Cognito as the primary authentication and authorization solution.
• Redesigned the Salesforce integration and underlying data model, improving data consistency and reliability between internal and external systems.
• Developed analytics and pricing interfaces used for fleet-risk and insurance-pricing workflows.
• Worked with product and business stakeholders to translate insurance workflows and operational requirements into technical solutions.
• Took a leadership role in a dedicated team of 3 developers, conducting cross code reviews of colleagues' work to maintain code quality and shared standards.

AI-assisted golf coaching platform connecting golfers with professional coaches, combining lesson management, player-performance analytics, progress tracking, and personalized AI-generated coaching recommendations.

• Built a RAG-based AI coaching assistant that generated personalized player feedback, progress summaries, and training recommendations using historical performance data, coaching notes, goals, and previous lessons.
• Implemented a retrieval pipeline using Amazon S3 vector buckets and Amazon OpenSearch to store and search embeddings generated from player and coaching data.
• Prepared coaching content and historical player records for retrieval by splitting source data into semantically meaningful chunks and attaching player, lesson, and content metadata.
• Grounded LLM responses in retrieved coaching and player data rather than relying exclusively on model knowledge.
• Added relevance filtering and fallback behavior so the assistant avoided generating recommendations when sufficiently relevant supporting data could not be retrieved.
• Included references to retrieved coaching records and source material in generated outputs, allowing coaches to review the information behind AI recommendations.
• Developed performance-analysis logic that transformed historical golf data into metrics and trends used by coaches to evaluate player development.
• Built interactive Chart.js dashboards for visualizing player progress and performance over time.
• Delivered coach-facing functionality for reviewing players, performance history, progress, and coaching recommendations.
• Built React features from Figma designs and integrated them with Python APIs.
• Integrated Auth0 authentication and added Cypress end-to-end coverage for critical golfer and coach workflows.

Open app framework enabling third-party payment services to integrate with a unified checkout and payments API. Merchants can rapidly add or connect PSPs, payment methods, fraud providers, and other payment-related services to their checkouts and payment flows.

• Rebuilt the connection layer, making the client SDK almost 2x faster.
• Added and improved payment methods including Google Pay, Apple Pay, Alipay, Checkout, Atome, PromptPay, and Apaya.
• Built stable end-to-end payment flows consistent across all platforms with backend, mobile, and QA teams.
• Improved overall performance and worked with product teams to make payment steps clearer and more reliable.
• Strengthened product quality with robust end-to-end tests and an improved testing process.

Enterprise platform for a leading travel company automating supplier-driven booking changes on a serverless AWS backend. The backend is built on AWS Lambda, API Gateway, and DynamoDB, while the React-based UI gives operations teams tools to review and initiate booking updates. The system follows modern IaC and CI/CD practices for scalability and fault tolerance.

• Designed serverless backend services on AWS Lambda, API Gateway, and DynamoDB to automate supplier-driven booking changes.
• Developed React and TypeScript components for the operations dashboard.
• Built and managed infrastructure as code with Terraform for consistent, scalable environments.
• Created GitLab CI/CD pipelines for automated builds, testing, and multi-environment deployments.
• Integrated Datadog monitoring and observability, improving incident response.
• Conducted code reviews, authored technical documentation, and mentored new team members.

Large-scale retail platform migrated from a monolith to serverless microservices on AWS. Services were built in Python on AWS Lambda, API Gateway, DynamoDB, and SQS, with infrastructure managed via AWS CDK, integrations with multiple third-party systems, and optimized communication between distributed components.

• Designed and implemented serverless microservices with Python, AWS Lambda, API Gateway, DynamoDB, and SQS.
• Maintained infrastructure as code with AWS CDK for reproducible environments.
• Built integrations with cloud e-commerce, payment systems, geo-location services, SAP ERP, and more.
• Built Azure DevOps CI/CD pipelines, reducing delivery time.
• Implemented comprehensive unit tests and conducted code reviews to maintain team standards.
• Onboarded and mentored new team members.

Web & software development company delivering projects across multiple industries, where Yura focused primarily on fintech and healthcare:

1. Payment platform with multi-currency wallets and PCI DSS/GDPR compliance.
2. Cloud healthcare platform for patient management and scheduling.

• Developed secure, scalable platforms for financial and healthcare applications.
• Implemented payment processing, account management, appointment scheduling, and reporting workflows.
• Ensured data integrity, security, and regulatory compliance across all services.
• Monitored system performance and applied optimizations to improve reliability.
• Contributed to code quality through unit testing and code reviews.

Cloud-based e-commerce platform that allows users to buy or rent vehicles online with secure payment processing, dynamic pricing, and real-time vehicle availability management. The platform connects dealerships, private sellers, and customers through a unified digital experience.

Key features: users can browse cars, motorcycles, and vans with advanced search, filtering, and comparison tools, support both one-time vehicle purchases and flexible rental plans, integrated with major payment gateways (Stripe) for deposits, full payments, and refunds, real-time reservation management with automatic calendar synchronization and notifications, sellers can manage inventory, pricing, availability, and view analytics through an intuitive dashboard.

• Designed and developed the platform with Python, Django REST, and AWS EC2.
• Integrated Stripe for deposits, full payments, refunds, and secure transactions.
• Implemented frontend features and fixed bugs to improve user experience.
• Ensured performance, reliability, and data consistency across backend and frontend.