Michael James

• Real-Time Predictive Maintenance System: Developed an edge-based predictive maintenance system for industrial machinery using Azure, Kubernetes, ClearBlade IoT, and Apache Kafka for real-time data streaming. Integrated Docker for containerization to enable scalable deployments across edge devices. Leveraged CI/CD pipelines and Azure Machine Learning to automate model updates and forecasting, enhancing decision-making processes for predictive maintenance.
• Intelligent Video Analytics for Smart Cities: Built a real-time traffic monitoring system using GCP, Kubernetes, ClearBlade Edge AI, and Kafka to manage backend data streaming. Deployed predictive models using PyTorch, NumPy, and Pandas, utilizing Docker and GCP Cloud Functions for containerized environments. Ensured real-time analytics and seamless communication across components through inter-process communication (IPC) and continuous CI/CD integration.
• IoT-based Asset Tracking Platform: Designed an IoT-based asset tracking platform with ClearBlade IoT, Azure, PostgreSQL, MongoDB, and Apache Kafka to manage device data efficiently and perform historical analytics. Integrated Docker for containerization and automated deployment through CI/CD pipelines. Incorporated LLM-based microservices to handle natural language queries for asset insights, improving accessibility and decision-making. Led cross-functional collaboration using Confluence and Jira, optimizing portfolio management processes through enhanced communication skills, product management best practices, and continuous refactoring of microservices using Flask and Git.

• Predictive Analytics for Customer Segmentation: Developed deep learning models for customer behavior prediction and segmentation using MongoDB, and integrated them with REST APIs for real-time predictive models. Scaled models and APIs with Kubernetes and Docker, enabling efficient forecasting and decision-making for business objectives. Managed workflows and cross-functional communication using Jira, Confluence, and Git.
• Fraud Detection System for Financial Transactions: Built and deployed a real-time fraud detection system with MongoDB and Kafka for data streaming, utilizing Kubernetes and Docker for scalable deployment. Enhanced detection accuracy through advanced data science methodologies (NumPy, Pandas) and model refactoring using FastAPI. Acted as a mentor and coach for junior engineers, fostering communication skills and driving collaborative success across the team.

• Data Reporting & Optimization: Achieved a 3% improvement in reporting accuracy by developing advanced tools using PyTorch, integrated with Apache Kafka and AWS Lambda, reducing machine setup time by 18%.
• Server Configuration Management: Developed backend configuration management tools to track server settings and created monitoring reports using AWS Lambda, Apache Spark, and Excel, enabling real-time data refresh every second through in-memory caching.
• Resource Utilization Optimization: Streamlined customer resource utilization by developing backend systems for NLP-based insights, reducing turnaround time by 15%, addressing customer needs, and enhancing cross-functional collaboration.