• Managed live payment workflows for enterprise clients, processing multi-currency transactions, refunds, chargebacks, and scheduled payouts using Python (FastAPI, Django REST) on a daily basis.

• Monitored Kafka payment event streams in real-time, reacting instantly to failed transactions, cancellations, and disputes, improving event reliability to 98% and reducing processing delays by 20%.

• Automated retry logic and background jobs with Celery and RabbitMQ, ensuring payment retries, client notifications, and financial operations ran smoothly, increasing retry success by 15%.

• Built real-time analytics pipelines with Spark, PySpark, Kafka Streams, and Redis, generating live KPIs like transaction success rates,

revenue trends, ARPU, and payment efficiency, helping teams take immediate action.

• Managed PostgreSQL for transactional data and MongoDB for dynamic client metadata, updating rules and discount configurations in response to promotions or changing business requirements.

• Designed and maintained RESTful and GraphQL APIs (FastAPI, Ariadne) for internal dashboards, operational tools, and client portals,

providing live transaction and invoice data to support finance and operations.

• Created interactive dashboards using React.js, TypeScript, Chart.js, and Material UI to visualize failed payments, refunds, revenue alerts, and transaction trends, allowing operations teams to detect and resolve issues in real-time.

• Developed microservices and backend tools in Python and Node.js for event validation, financial exports, and automated reconciliation, improving day-to-day operational efficiency.

• Packaged services with Docker, deployed on AWS EKS using Helm charts and blue-green deployments, while managing serverless

workflows with AWS Lambda, SQS, S3, and CloudWatch for continuous monitoring and backups.

• Developed a Python-based backend platform using Flask and FastAPI to handle millions of ad impressions and user interactions daily, delivering real-time engagement insights, content recommendations, and ad optimization metrics for Meta.

• Built modular microservices for user segmentation, ad scoring, personalized recommendations, and revenue tracking, exposing secure REST APIs for internal analytics and ad operations teams.

• Implemented low-latency, event-driven pipelines with Apache Kafka to stream user clicks, ad impressions, session events, and engagement metrics in real-time, powering instant predictions and content ranking.

• Created interactive dashboards with React.js, TypeScript, Redux, HTML5, and CSS3 to visualize live engagement trends, revenue anomalies, and priority alerts, enabling faster decision-making for ad optimization and campaign adjustments.

• Developed machine learning models using scikit-learn, TensorFlow, and Pandas for real-time engagement prediction, churn analysis, and personalized content ranking, improving ad targeting accuracy by 20% and user retention by 15%.

• Managed data storage and retrieval with PostgreSQL for structured data, MongoDB for unstructured logs, and Redis for caching highdemand predictions, reducing dashboard load times by 35%.

• Built asynchronous ML inference pipelines using FastAPI and asyncio to handle thousands of prediction requests per second, ensuring reliability during peak traffic.

• Developed anomaly and fraud detection services to identify unusual engagement patterns, click spamming, and bot activity, reducing false positives by 25% and strengthening revenue integrity.

• Deployed services on AWS (EC2, S3, Lambda, RDS), containerized with Docker, and orchestrated using Kubernetes (EKS) for high availability, auto-scaling production workloads.

• Automated CI/CD pipelines using GitHub Actions, Helm, and Terraform for continuous integration and secure deployments across development.