Bhargavi Karuku

About

AI/LLM engineer with production experience building an LLM-powered investment recommendation system using RAG and chatbots, deployed via Docker/CI/CD and scaled on Kubernetes. Demonstrated measurable performance wins (sub-200ms latency) through QLoRA fine-tuning and TensorRT INT8/INT4 quantization, plus strong MLOps/orchestration background (Airflow ETL + scoring, MLflow monitoring) and stakeholder-facing delivery using demos and Tableau dashboards.

Experience

Education

Key Strengths

• Built and deployed a production LLM/RAG investment recommendation system (Docker + CI/CD on cloud)
• Reduced inference latency to <200ms and lowered cloud costs via TensorRT INT8/INT4 quantization (2–4x GPU speedup; 50–75% memory reduction)
• Cut fine-tuning/training cost ~80% using QLoRA adapters while improving domain accuracy
• Strong orchestration experience: Airflow for daily ETL + scoring pipelines and Kubernetes for scaling LLM services
• Improved pipeline reliability with Airflow retries and Slack alerting to prevent delays in A/B testing and keep recommendations fresh
• Structured agent/workflow evaluation: unit/integration tests, synthetic data, A/B experiments, red teaming, and production monitoring with MLflow
• Effective cross-functional collaboration: translated RAG/LLM concepts into demos and Tableau metrics dashboards; added explainability and compliance checks based on PM feedback