Varun Duggirala

About Varun Duggirala

Full Stack AI Engineer with 4+ years of experience building scalable AI platforms…

Experience

1. Ai Software Developer

   Morgan Stanley

   Jul 2025 — Present · US

   Built and enhanced Python-based backend services for AI-driven internal applications supporting financial research and document intelligence workflows, reducing response latency by 30% across high-usage internal requests.Developed Retrieval-Augmented Generation pipelines integrating large language models, vector databases, and structured financial datasets, improving answer relevance and retrieval accuracy by ~35% for analyst-facing use cases.Designed agent-based orchestration workflows to coordinate API calls, retrieval layers, and reasoning steps for multi-stage tasks, reducing manual processing effort by 30% in recurring operational workflows.Deployed containerized AI inference services using Docker and Kubernetes with production-grade logging, latency monitoring, and service health controls, sustaining 99.5% uptime across production environments.Supported MLOps workflows using MLflow and Kubeflow for model version tracking, deployment consistency, and controlled release cycles across internal AI services.Built batch and near real-time data pipelines using SQL and Pandas to prepare transaction and reference data for model inference, improving downstream data readiness by 25%.Contributed to Edge AI experimentation for lightweight inference scenarios by optimizing selected model components for lower-memory deployment and reduced inference overhead in controlled internal environments.