Trajectory is building a continual learning platform to close AI agents' 'experience gap'
“These models are smarter and smarter but always feels like when you're talking to them, it's their first day on the job.”
agent-memory
16 tracked signals on agent-memory.
LLM synthesis destroys data provenance trails, creating critical risks in high-stakes agent applications
“Synthesis often destroys the paper trail of how these outputs were originated.”
LangSmith Engine and Context Hub turn agent traces into durable, versioned long-term memory for continual learning.
“Most agents don't learn, they just leave traces.”
Microsoft Foundry Agent Service adds procedural memory so agents reliably reuse proven execution patterns in production.
“agents often know the right facts and still fail the task because they do not execute the right procedure”
Microsoft Foundry separates agent identity, conversation state, and executable workspace into distinct controls
“Agent memories becoming ordinary cloud data architecture.”
DeepLearning.AI launches course teaching agents to learn from past sessions via behavioral and knowledge adaptation
“they can run into same problems again and again because they don't remember what they learned yesterday”
Graph-based memory outperforms flat markdown files for persistent AI agent recall
TurboQuant compresses agent retrieval embeddings to 3-4 bits, cutting memory cost 5x without degrading search quality.
“Today, we will see how you can cut memory cost of agent retrieval five times without breaking your search.”
NVIDIA NeMo Agent Toolkit now supports Amazon S3 Vectors as a custom persistent memory backend
Agent long-term memory requires persistent storage beyond stateless LLM conversational context
“LLMs are stateless machines.”
Apple explains agents learn via notes and tools, and Xcode's agent MD file is injected into every query so keep it short.
“You wouldn't want to put like your entire style guide in there every single time. You'll be eating up a lot of tokens.”
Twilio and Azure AI Foundry integrate to build omnichannel AI agents with persistent cross-channel memory.
“56% of leaders say they're getting nothing out of AI”
TiDB consolidates agent memory into one SQL database, replacing the three-database-plus-ETL stack most teams use.
“you're running millions of agents at the same time and each with their own state”
Microsoft demo shows SQL Server as persistent memory backend for AI agents in e-commerce
“I think agents can help by creating a recommendation system that knows who you are. It knows what you like, what you don't like.”
Neo4j engineer builds offline personal AI agent with graph-based memory on Raspberry Pi
Microsoft Foundry lets developers create multimodal AI agents that behave like people in a few clicks.
“Next year you're going to come and you're going to see agents that are no longer just chat bots. You're going to see agents that are like people.”