Mallikarjuna Reddy Gayam

About

ML/LLM engineer who has shipped production RAG systems (LangChain + HF Transformers + FAISS) with hybrid retrieval and cross-encoder re-ranking, deployed via FastAPI/Docker/Kubernetes and monitored with MLflow. Also partnered with wealth advisors at Edward Jones to deliver a client retention model with SHAP-driven explanations and a dashboard that improved trust, adoption, and reduced high-value client churn.

Experience

Education

Key Strengths

• Built and deployed enterprise RAG knowledge assistant reducing employee time spent searching policy documents
• Deep retrieval optimization: chunking strategy tuning, hybrid BM25+dense retrieval, cross-encoder re-ranking with latency tradeoffs
• Production ML deployment and ops: FastAPI + Docker + Kubernetes with monitoring and reliability practices
• Experimentation and evaluation rigor: labeled datasets, precision/recall measurement, baseline comparisons
• Robust ML pipeline orchestration and governance using Airflow + MLflow, including drift detection and retraining triggers
• Strong stakeholder collaboration and explainability: translated advisor needs into features and delivered SHAP-based, business-friendly dashboard that drove adoption and reduced churn