The Agentic Mind

Shuva Jyoti Kar is a Principal Engineer at Palo Alto Networks, where he architects next-generation firewall platforms designed to deliver complete visibility, control, and protection for enterprise AI applications, models, and LLMs. Prior to this, he served as a Senior Technical Leader at Cisco Systems, building high-scale data platforms for AI-driven network analytics. With a career rooted in the rigors of distributed systems, he was an early open-source pioneer as a committer and maintainer for the OpenDaylight OpenFlow Plugin. Having witnessed the pitfalls of fragile internal agent frameworks that break under prompt bloat or drift, he specializes in the "hard "side of AI. He brings a strict systems-engineering discipline to probabilistic computing, focusing on moving beyond prototypes to build scalable, serverless, and distributed autonomous systems that are as reliable and secure as the networks they manage.

Part I. The Foundations of Machine Memory.- 1. The Statelessness Crisis.- 2. The Architecture of Recall.- Part II. The Technical Persistent Layer.- 3. Advanced RAG: Beyond Semantic Search.- 4. Relational Memory with Knowledge Graphs.- 5. Virtual Memory & Paging (The MemGPT Pattern).- Part III. The Learning Loop & Evolution.- 6. Reflection & Self-Correction.- 7. Personalization & User Modeling.- Evaluation & Observability.- 8. The Art of Forgetting (Garbage Collection).- 9. Multi-Agent Shared Memory.- 10. Evaluation & Observability.