Anvith Reddy T

About

ML/AI engineer with hands-on ownership of fraud detection and investigator-assist systems, combining anomaly detection with RAG-based LLM summarization in production. Stands out for translating research ideas into reliable cloud-deployed workflows that improved precision to 92%, cut review time by 25-30%, and increased investigator throughput by roughly 30% while also building reusable Python infrastructure for team-wide velocity.

Key Strengths

• Owned end-to-end production ML/AI systems from experimentation through deployment and monitoring
• Strong fraud detection and anomaly detection expertise
• Shipped RAG-based LLM summarization and Q&A workflows for investigators
• Balanced latency, cost, quality, and safety in production GenAI systems
• Built cloud-deployed ML pipelines with Docker, AWS, Azure, and CI/CD
• Implemented robust model and system monitoring including precision, recall, drift, latency, and job success rates
• Applied explainability techniques with SHAP and LIME to improve user trust
• Improved fraud model performance to 92% precision and 87% recall
• Drove 25-30% reduction in case review time
• Increased investigator throughput by about 30% without additional headcount
• Built reusable Python service/library architecture that reduced new workflow setup from 7 days to 1-2 days
• Improved adoption and trust by converting LLM outputs into structured, evidence-grounded summaries

Experience

Education

Skills

Languages

Certifications

Publications

1 publication