Articles
Technical deep dives on RAG, LLM systems, AI architecture, and production deployment. Published on Medium and Dev.to.
Spec-Driven Development: Moving AI Coding from Experimentation to Production Discipline
Why AI coding needs a contract-first workflow with clear specifications, planning gates, traceable implementation, and verification before production use.
Read on Website →RAG Patterns: A Practical Guide to Retrieval Architectures
An expanded practical map of modern RAG patterns: classic, hybrid, reranking, graph, temporal, security-aware, tool-augmented, agentic, and production layering strategies.
Read on Website →Kubernetes Centralized Logging: Best-in-Class Observability Pipeline for 100+ Microservices
How to design a resilient Kubernetes logging architecture with OpenTelemetry Collector, Kafka, Vector, OpenSearch, Grafana, ElastAlert 2, and S3 for 100+ microservices.
Read on Website →Building Production RAG: Cost Optimization Strategies
How to reduce RAG pipeline costs by 66% without sacrificing quality. Practical strategies: prompt caching, embedding model selection, batch processing, and cost-per-query optimization. Includes real numbers from production deployment.
Read on Medium →Evaluating RAG Systems: Beyond Automated Metrics
Why ROUGE and BERTScore fail for RAG evaluation. The case for human-in-the-loop validation, building labeled datasets, and continuous monitoring. Covers uncertainty sampling and production evaluation strategies.
Read on Dev.to →Conversational AI in Healthcare: Domain-Specific Challenges
Lessons from building health insurance chatbots. PII handling (HIPAA), regulatory compliance, conservative response strategies, and maintaining accuracy in regulated domains. Real examples from production deployment.
Read on Medium →Architecture Lessons from 20 Years: Systems Thinking at Scale
Evolutionary lessons building systems from monoliths to microservices to serverless. Why governance matters. Why cost is architecture. Why observability is a first-class concern. Examples from Yale and enterprise systems.
Read on Dev.to →Hybrid Retrieval for RAG: Semantic + BM25 Search
Why semantic search alone isn't enough for domain-specific RAG. How to combine vector similarity with keyword matching (BM25) using reciprocal rank fusion. Benchmark results showing 8% accuracy improvement.
Read on Medium →LLM Prompt Engineering: Best Practices for Production
Techniques that work in production: few-shot learning, chain-of-thought prompting, output formatting, and temperature tuning. How to design prompts for consistency and cost optimization. Examples from health insurance domain.
Read on Dev.to →Subscribe & Follow
Get notified when new articles are published.
- 📰 Medium: Follow @antonglenbovitch
- 💻 Dev.to: Follow @antonglenbovitch
- 🐙 GitHub: Follow AntonGlenbovitch
- 🔗 LinkedIn: Connect with Anton