Designing Context for Tasking Agents
Context assembly for agents that research, delegate, and execute.
Blog
Designing a Memory Layer for GenAI Chat Applications
Memory, retrieval, and context assembly for LLM products.
Conversational Context and Agent Context Are Different Problems
Conversational continuity and agentic execution are different problems. They should not share the same memory assumptions.
Scaling GenAI Chatbots Is Cheap. Scaling Agents Is Expensive.
Generative AI chatbots are affordable to run at scale. Tasking agents are not. The gap is larger than the market language suggests.
There Is No Universal Memory System for AI Products
Every AI product has to manage context. The right memory design depends on what the system is trying to do.