Uploaded August 2026 | Updated September 2026, 2 weeks ago
Want an AI agent that remembers project context instead of starting every conversation from zero? In this demo, we configure agent memory and build a meeting-transcript summarizer that carries forward decisions, risks, owners, and reporting preferences across conversations.
Most AI agents start a conversation from zero. Agent Memory gives the agent continuity: it retains relevant information from earlier conversations and uses it when it helps the same user in a future chat. In the example, a project manager's reporting preferences, stakeholder roles, launch decisions, risks, and follow-up items are used to create more personalized and useful executive summaries over time.
Chapters:
0:00 Introduction to agent memory
0:54 Configure agent memory
1:26 Build the summarizer workflow
2:20 Enable memory across chat threads
2:45 Add durable project context
3:30 Save the first meeting summary
5:25 Use memory in a new chat
6:15 Maintain continuity and manage memory
What this demo covers:
- Configuring an embedding model for Agent Memory
- Enabling memory for an Agent Builder workflow
- Sharing memory across chat threads
- Building a transcript-summarizer workflow
- Capturing durable project context, roles, decisions, and risks
- Using prior context in a new chat without mixing it with current decisions
- Tracking completed and unresolved follow-ups across meetings
- Inspecting retrieved and stored memory
- Deleting individual memory records when they are no longer needed
- Understanding user- and workflow-scoped memory controls
With Agent Memory, a transcript summarizer stops being a one-time meeting tool and becomes a continuity assistant that understands the project, tracks relevant follow-up, and delivers summaries in the format the user needs.
Related documents:
Agent Builder - Oracle AI Database Private Agent Factory 26.7: docs.oracle.com/en/database/oracle/agent-factory/26.7/paias/agent-builder.html
Configure Agent Memory: docs.oracle.com/en/database/oracle/agent-factory/26.7/paias/configure-agent-memory.html
Oracle AI Database Private Agent Factory documentation: docs.oracle.com/en/database/oracle/agent-factory/26.7/paias/introduction.html
#codeinnovate #AIAgents #AgentMemory #GenerativeAI #EnterpriseAI
Want an AI agent that remembers project context instead of starting every conversation from zero? In this demo, we configure agent memory and build a meeting-transcript summarizer that carries forward decisions, risks, owners, and reporting preferences across conversations.
Most AI agents start a conversation from zero. Agent Memory gives the agent continuity: it retains relevant information from earlier conversations and uses it when it helps the same user in a future chat. In the example, a project manager's reporting preferences, stakeholder roles, launch decisions, risks, and follow-up items are used to create more personalized and useful executive summaries over time.
Chapters:
0:00 Introduction to agent memory
0:54 Configure agent memory
1:26 Build the summarizer workflow
2:20 Enable memory across chat threads
2:45 Add durable project context
3:30 Save the first meeting summary
5:25 Use memory in a new chat
6:15 Maintain continuity and manage memory
What this demo covers:
- Configuring an embedding model for Agent Memory
- Enabling memory for an Agent Builder workflow
- Sharing memory across chat threads
- Building a transcript-summarizer workflow
- Capturing durable project context, roles, decisions, and risks
- Using prior context in a new chat without mixing it with current decisions
- Tracking completed and unresolved follow-ups across meetings
- Inspecting retrieved and stored memory
- Deleting individual memory records when they are no longer needed
- Understanding user- and workflow-scoped memory controls
With Agent Memory, a transcript summarizer stops being a one-time meeting tool and becomes a continuity assistant that understands the project, tracks relevant follow-up, and delivers summaries in the format the user needs.
Related documents:
Agent Builder - Oracle AI Database Private Agent Factory 26.7: docs.oracle.com/en/database/oracle/agent-factory/26.7/paias/agent-builder.html
Configure Agent Memory: docs.oracle.com/en/database/oracle/agent-factory/26.7/paias/configure-agent-memory.html
Oracle AI Database Private Agent Factory documentation: docs.oracle.com/en/database/oracle/agent-factory/26.7/paias/introduction.html
#codeinnovate #AIAgents #AgentMemory #GenerativeAI #EnterpriseAI
