Knowledge & Memory Articles
23 articles · Page 1 of 2
Build Context Editing Into Your Agent and Cut Token Use 84%
Stale tool results pile up in long conversations, raising cost and hurting accuracy. Context editing and the memory tool fix both. Here's how to build them.
How to Build a Forgetting Policy for Agent Memory
An agent that remembers everything eventually surfaces the wrong fact. Here's how decay, eviction, and supersession keep memory useful instead of just large.
Your agent doesn't need more memory. It needs a graph.
Session memory forgets. RAG retrieves but doesn't reason. Context graphs give your CX agent a persistent, queryable model of every customer, issue, and relationship it's ever encountered -- at a fraction of the token cost.
When vector search fails: GraphRAG for complex CX agent queries
Flat vector search breaks on multi-hop CX queries like 'what discount applies to premium customers in California who signed up before 2024?' GraphRAG uses a knowledge graph to answer these correctly.
Agent warmup: preload customer context before the first word
Most agents ask for information they already have. Here's how to preload the right customer context before an interaction starts, so your agent sounds like it knows the customer from word one.
What your agent does when it knows nothing about a customer
The cold-start problem hits every CX agent on first contact: no history, no profile, no context. Here's a practical architecture for handling first-contact customers without making them feel like strangers.
Agents that learn while they sleep: async memory consolidation
Post-session async consolidation is the missing layer in most agent memory stacks. Here's how Anthropic's Dreaming primitive works, why Harvey saw a 6x jump, and how to implement the same pattern without Anthropic's API.
Your Agent's Context Window Is RAM, Not Storage
Most agent failures trace back to one mistake: treating the context window like a database. Here's the RAM model that fixes attention dilution, latency spikes, and ballooning costs.
How to Build a Tier-1 Chat Agent That Resolves (Not Just Deflects)
Bots claim 40% deflection; re-contact data says half is fake. Build the architecture that cuts tickets: auth-gated KB, calibrated confidence, escalation with context.
Your Agent Completed the Task. It Also Forgot 87% of What It Knew.
Task completion hides a silent failure: agents forget 87% of stored knowledge under complexity. New research reveals why standard evals miss this entirely.
The RAG You Built Last Year Is Already Outdated
RAG has branched into 5 distinct architectures: Self-RAG, Corrective RAG, Adaptive RAG, GraphRAG, and Agentic RAG. Here's when to use each and how to choose.
Your RAG Returns Wrong Answers. Upgrading the Model Won't Help
Most RAG quality problems are retrieval problems, not model problems. Bad chunking, wrong embeddings, and missing re-ranking cause more hallucinations than model capability gaps.
El briefing de Signal
Un email por semana. Cómo los equipos líderes de CS, ingresos e IA están convirtiendo conversaciones en decisiones. Benchmarks, playbooks y lo que funciona en producción.
