Digital Transformation Insights
Expert perspectives on legacy modernization, AI transformation, and enterprise engineering best practices.
How we implemented a comprehensive memory system that gives both AI agents and our Orchestrator persistent memory across conversations, featuring a 4-tier hybrid architecture with semantic search, automatic episode creation, and intelligent context injection.
How we reduced token usage by 85% while maintaining intelligent tool selection across hundreds of MCP servers, agents, and integrations using a hybrid BM25 + semantic search approach.
The story of building a production-grade AI orchestrator powered by Temporal that coordinates multiple agents, learns from experience, and doesn't fall apart when things go wrong. With 99.7% completion rates and real customer results.