• Architected microservices advertising platform using Node.js and TypeScript, processing 500K+ daily API requests with 99.9% uptime and <100ms response time through Redis caching

• Designed RESTful APIs with OpenAPI specs enabling third-party integrations and partner onboarding, reducing integration time by 60%

• Built campaign management dashboard with React and TypeScript featuring real-time analytics, optimization tools, and dynamic bidding strategy adjustments

• Containerized microservices with Docker and deployed to AWS ECS with auto-scaling, achieving 70% faster deployments and zero-downtime updates across 8+ services

• Implemented OAuth 2.0 and JWT authentication with Auth0, ensuring GDPR compliance and PII protection for 100K+ monthly users with zero security incidents

• Developed and optimized the website’s search function source code using MERN stack technologies, resulting in a 15% increase in user

engagement by enhancing search accuracy and reducing response time by 10%.

• Designed and implemented RESTful APIs for backend services, supporting over 5,000 monthly active users, ensuring a 98% uptime and

improving data retrieval speed by 12%.

• Collaborated with a small cross-functional web application development team to design a conversational AI chatbot, enhancing user

experience and contributing to a 10% increase in user interaction during the early launch phase.

• Implemented Apache Kafka data pipeline processing 50K+ events/minute with exactly-once semantics to power analytics, audience targeting, and user segmentation

• Built Java Spring Boot content platform serving 2M+ monthly page views, reducing server costs 30% through CDN caching and query optimization

• Designed RESTful microservices API handling 100K+ daily requests across PostgreSQL and MongoDB, improving response time 45% and reducing database load 35%

• Developed AI recommendation engine using Python and OpenAI GPT-4 API, increasing user engagement 35% and conversion rates 18% through personalized suggestions

• Collaborated with a team of developers on fine-tuning GPT-4 and LlamA-2 Large Language Models (LLMs) to optimize the Ruby code generation process for Trimble - SketchUp, resulting in a 20% increase in accuracy and relevance of generated code.

• Implemented a scalable search function using Microsoft Azure Functions and Langchain for the Retrieval Augmented Generation (RAG) pipeline, leading to a 30% improvement in efficiency and performance.

• Facilitated regular sprint reviews as the project lead with the team to track progress, address challenges, and ensure alignment with project goals, ultimately resulting in an on-time delivery of the Copilot extension to Trimble - SketchUp.

• Engineered scalable Python backend services for SketchUp’s AI-powered 3D generation platform, processing 10K+ daily API requests with average latency <200ms through asynchronous processing and optimized Azure infrastructure

• Built automated code generation pipeline integrating Azure Cognitive Services and OpenAI GPT models, fine-tuning embeddings and implementing RAG (Retrieval-Augmented Generation) to improve output accuracy by 40%

• Architected serverless data processing pipeline using Azure Functions and Pinecone Vector DB, implementing distributed caching and retry mechanisms to process 1M+ embeddings daily with 99.8% success rate

• Partnered with cross-functional ML and product teams to deploy 5+ production models using Azure ML pipelines, implementing A/B testing framework and monitoring dashboards that enabled data-driven optimization decisions