Senior Software Engineer (AI / Machine Learning) Oscilar
• Designed and developed backend services using Java (Spring Boot), Python for Oscilar AI Risk Decisioning Platform, enabling real-time fraud, credit, and compliance decisioning across financial transactions.
• Built low-latency decisioning services, reducing average response time to under 100ms for transaction risk evaluation across high-volume streaming pipelines.
• Implemented event-driven architecture using Kafka, processing thousands of events per second for real-time ingestion and decisioning workflows.
• Designed and implemented RESTful APIs for transaction processing and risk evaluation, enabling reliable communication between distributed microservices.
• Designed and optimized PostgreSQL schemas for transaction and user data, improving query performance and ensuring consistency for high-volume financial workloads.
• Implemented caching strategies using Redis, reducing data retrieval latency by ~40% and improving performance for real-time decisioning services.
• Developed Python-based services for machine learning and AI workflows using FastAPI, enabling scalable and low-latency model integration.
• Integrated machine learning models using PyTorch, XGBoost, and Scikit-learn, improving fraud detection accuracy and reducing false positives by ~20%.
• Designed microservices-based architecture deployed on Kubernetes, improving scalability, resilience, and supporting high availability across distributed services.
• Built internal dashboards using React and TypeScript, improving analyst efficiency and reducing manual review time for flagged transactions.
• Developed LLM-powered features for case summarization and risk analysis, improving investigation efficiency and reducing manual effort.
• Built retrieval-based pipelines combining real-time and historical data, improving contextual accuracy and explainability of AI-generated insights.
• Leveraged AWS infrastructure for scalable deployment, monitoring, and high availability of real-time decisioning systems.
• Implemented automated testing and CI/CD pipelines, reducing deployment time by ~35% and improving release stability across backend services.
• Leveraged AI-assisted development tools to improve code quality and reduce debugging time across backend and frontend workflows