AI Agent Memory: Chat History & Semantic Caching | Lino Tadros | Azure Cosmos DB Conf 2026

AI agents are only as intelligent as their ability to remember. Without persistent memory, every conversation starts from scratch — costing tokens, increasing latency, and delivering disconnected experiences. Lino Tadros (Microsoft Regional Director, Founder & Principal Architect at The Training Boss) shows how to use Azure Cosmos DB as the unified memory layer for agentic AI applications. You'll build two essential containers from scratch: • Chat History — preserves multi-turn conversations across sessions and devices • Semantic Cache — uses vector search to return prior LLM completions for semantically similar prompts, avoiding redundant API calls See the data modeling decisions that matter: partition key strategy for multi-tenant workloads, document design that balances write distribution with query locality, vector indexes tuned for fast similarity search, and TTL policies for automatic lifecycle management. Walk away with working code patterns you can apply immediately. 👤 Connect with Lino Tadros 📝 Distinguished executive leader and renowned technical expert in AI, Machine Learning, and IoT. Leads cross-functional architectural teams to award-winning performance by developing strategic roadmaps and powering enterprise-wide projects. Serves as board member and advisor for multiple corporations delivering strategic guidance on product line developments and business solutions. Industry influencer and mastermind of strategic programs and innovations leading modernization efforts to alter the global IT landscape as Microsoft Regional Director. Partnered with Microsoft to consult major corporations on Azure integrations; trained over 1,000 global employees and architects across US, Canada, Europe, Middle East, and Australia. Invited into elite Microsoft Regional Director program as Top 1% of global SMEs—maintains direct line of communication to Microsoft executive leaders and Office of the President for Technical and Business Influence. Piloted multi-million