Swetha Selvaraj

• Designed and deployed AI pipelines for enterprise observability and HRMS use cases improving retrieval accuracy, cost usage into account and reducing response latency via prompt optimizing techniques, reranking, and strict metadata filtering.
• Built AI microservices in Python and .NET integrating AWS Bedrock and Gemini models with function calling, achieving consistent structured outputs for enterprise workflows.
• Automated system metrics and ITSM data pipelines using MongoDB, ServiceNow (CMDB, Incident, Problem, Change), enabling real-time dashboards and reducing manual data reconciliation effort.
• Contributed to code reviews, handled client interactions, and been the primary POC for requirement gathering and delivery.

• Developed QA chatbot for Med research papers using RAG and OpenAI API served using FastAPI with Gradio UI.
• Implemented text-to-speech services for learning platforms using Azure AI Voice TTS models.
• Developed and evaluated ML based anomaly detection and forecasting models on system logs and infrastructure metrics,
• performing feature engineering and iterative experimentation to improve detection accuracy.

Developed a skincare recommendation engine using collaborative filtering and regression models.

Improved recommendation accuracy via feature engineering on customer questionnaire data.