Anton Glenbovitch

Anton Glenbovitch is a software engineer with 20+ years of experience building enterprise systems, currently focused on production-grade AI applications including Retrieval-Augmented Generation (RAG), LLM orchestration, and scalable AI infrastructure on AWS.

• Production-grade RAG pipelines and vector search systems
• LLM orchestration and agent-based workflows
• AI system architecture and cloud deployment (AWS)

Enterprise Claim AI Platform

A production-style AI system designed to analyze insurance claims using RAG architecture, vector search, and LLM reasoning.

View Project on GitHub →

In production AI systems, the primary challenges are not model selection, but data quality, retrieval accuracy, and system design. Reliable systems require strong retrieval pipelines, clear guardrails, and continuous evaluation to minimize hallucinations and ensure consistency.

These are the core references and tools I rely on when designing and building production-grade RAG systems.

• AWS Bedrock – foundation model orchestration layer for building scalable AI systems on AWS.
• OpenAI Embeddings – practical reference for semantic search and vector-based retrieval pipelines.
• Retrieval-Augmented Generation (RAG) – foundational architecture combining retrieval with LLM reasoning.
• OpenSearch Vector Search – scalable indexing and retrieval strategies for production RAG systems.
• Constitutional AI (Anthropic) – techniques for improving LLM alignment and reducing hallucinations.

15+ years building enterprise systems at Yale University and as an independent consultant, working across backend systems, data platforms, and AI-driven applications.

• GitHub
• LinkedIn