Floki: Building Agentic Workflows with Dapr
summary
Floki is an open source Python framework that layers LLM-savvy tasks and agents on top of Dapr Workflows, turning function calls into orchestrated agentic pipelines. It demonstrates using function-calling with OpenAI models, a REACT-style loop, and an orchestrator that picks which agent to run (example: a secret-holder agent calls a tool and returns the secret). Examples show reduced boilerplate (task wrappers replace verbose activity code), integration with Redis/Dapr Store for memory, and compatibility with Autogen-style multi-agent patterns.
AI-written summary · about
chapters (11)
- 0:00Floki and agentic workflows
- 1:28GenAI in enterprise security
- 3:11LLM autonomous agents
- 6:03LLM tool integration
- 9:13ReAct pattern for agents
- 13:52Specializing agents
- 18:07Extending Dapr workflows
- 23:49Simplifying LLM integration
- 27:05Multi-agent workflows
- 32:30Floki and Dapr future
- 33:36Closing and resources
related talks
01h 09mFloki: Defining Agentic Workflows for Security
45mDapr Workflows: How to Write Reliable, Fault-Tolerant Applications...
39mDapr in 2025: Durable Execution for AgenticAI Systems and Becoming the Backbone of App Platforms
25mDapr Workflows: The Key to Simplifying Business Process Orchestration
01h 00mReliable Agentic Systems need Durable Execution 🤖
30mHarnessing Crossplane and Dapr for DevOps: FICO’s Platform Engineering Journey to Increase Velocity
46mDapr in Practice
22mBuilding Real-World Applications with Dapr: A Startup Developer's Journey