From the course: Agentic AI for Developers: Concepts and Application for Enterprises
Join today to access over 24,700 courses taught by industry experts.
Memory in agentic AI
From the course: Agentic AI for Developers: Concepts and Application for Enterprises
Memory in agentic AI
- [Instructor] The final component of agentic AI that we will discuss is memory. Memory plays a key role in agentic AI. It keeps track of the conversations and tasks happening in an agent. This provides context to the planner to plan the next steps. It also keeps the history of actions for future reference. There are typically two types of memory in agents, namely short-term memory, and long-term memory. Let's discuss short-term memory now. In general, the life of short-term memory is the execution of a single prompt or goal. It also can be a single chat session with a user with multiple goals. This provides context when calling the LLM for planning. It also gives the history of what has happened so the LLM can decide on the next action. It keeps track of history of the workflow being executed to achieve the goals. This can be LLM calls for planning and the LLM responses. It also contains all the tool calls and the results returned from the tools. Memory also contains the chat history…
Download courses and learn on the go
Watch courses on your mobile device without an internet connection. Download courses using your iOS or Android LinkedIn Learning app.
