Agents and Memory Architecture
summary
Memory becomes the production requirement: short‑term chat blobs persist to a session store and Agent Core auto‑extracts, consolidates and embeds into long‑term vectors with TTL and built‑in encryption, episodic records (context, goal, action, outcome, insight) support reflection and reduce redundant API calls. Implementation uses Amazon Bedrock AgentCore and selectable vector stores; one migration shows S3 tables 98% cost saving versus RDS and a country onboarding cut from 1–2 months to ~1 minute at ~€1, while operators target a 2,000× transaction throughput increase.
AI-written summary · about
chapters (6)
- 0:00Introduction and roadmap
- 1:06Evolution of agents
- 4:55Why memory matters
- 12:44FintruX modernization journey
- 23:13Agent Core memory deep dive
- 44:50Best practices and resources
related talks
18mArchitecting Agent Memory: Principles, Patterns, and Best Practices
54mThe Anatomy of Memory in Humans and AI Agents
25mWhat does it take to build long-term state for AI agents?
46mThe Anatomy of Memory in Humans and AI Agents
01h 49mBuilding Memory Augmented Agents with MemoRizz
14mMake Agents Remember the Right Context at the Right Time
28mAI Agents That Remember, Share, and Cost Less with Meko | Introducing Meko
28mAgentic Memory