Building AI Agents That Actually Remember

​AI agents can look great in a demo, then fall apart the moment memory, retrieval, and real-world complexity matter.

​This meetup is about what it actually takes to build agent systems that can remember the right things, retrieve the right context, and behave more reliably in production.

​We'll hear from speakers at Google, Tavily, and Oracle, each sharing practical lessons from building real AI systems, followed by audience Q&A and time to meet others working in the space.

​This is a relaxed, down-to-earth local chapter meetup designed for learning, good conversations, and connecting with other engineers and builders in NYC.

TOPICS WE'LL EXPLORE

​Some of the themes across the evening will include:

• ​Agent memory and long-term context

• ​Retrieval and knowledge systems

• ​Databases and infrastructure for AI agents

• ​Building agent systems that are more reliable in production

• ​The difference between demo-quality agents and production-ready systems

​AGENDA

​5:00 - 6:15 PM | Doors Open — Arrival & Networking
 Check-in, food, drinks, and informal networking
6:15 - 6:30 PM | Welcome & Kickoff
 Opening remarks from MLOps Community and Oracle
6:30 - 7:00 PM | Richmond Alake, Oracle
 Memory Engineering and the Rise of Memory-Aware Agents
7:00 - 7:30 PM | Dean Sacoransky, Tavily
 Context Management in Research Agents
7:30 - 8:00 PM | George Pearse, Visia
 Efficient Context Management for Computer Vision
8:00 - 8:30 PM | Pier Paolo Ippolito, Google
 Moving from Prompt to Production: A Standardized Lifecycle for AI Agents
8:30 - 9:30 PM | Open Q&A + Networking
 Audience questions, discussion, drinks, and networking

​WHO SHOULD ATTEND

​Engineers, ML practitioners, and technical builders interested in AI agents, retrieval systems, memory architectures, and real-world LLM infrastructure.