agentic sols
how the interact with each otheer
Agents and the key. directional markets are real. Me talking interaction with strategies we’ll discuss during this case. How much you can studies coming from the real world examples, marketing scenarios, sub percentage scenario will predict diamonds. These ones and we don’t actually need to be calling
chat experiences :
Reactive sols: Ask Approach : Chatbot,Rag,Copilot
Autonpmous sols: Do appraoch: agents can do
defs:
What is aagent? componet, that has tools (services internet, github), strategy (self critic, Chain of though, planning - prompt enginerring technique though steps, , how agents communicate with each other,), states (memory, context, microservice) , goals..
agents doing planning - support center
Agents : Actions:
Agent to agent interaction:
multi agents : can have overal goal and local goals, collaboration
Componding systems: of muli edgents, instead of doing massive prompt engineering.
Example: writer agent , reviewer agent each is tied to a limited scope.
business problem > identify personas (agents) > agentic workflow … or the otherway around to define the buisness process.
Agentic Architectures: vertical (reporting) or horizonal (feedback high) or mixed . THink of a buisness hierarchy and how comminication is done: task dependnat
In a vertical the primary agent can redirect to a lower level agent.
Orleans: distrubited .NET - dist apps, Grains = agents, Silos = distrubted work, Clusters = collections of silos,
“
Given a problem, descript the roles requires, then model the roles with agents.