Please enable JavaScript.
Coggle requires JavaScript to display documents.
Agentic AI - Coggle Diagram
Agentic AI
Notes
deterministic systems—such as Robotic Process Automation (RPA)—that follow fixed rules and predefined outcomes
RPA remains critical for highly compliant, secure and resilient operations.
GenAI focuses on creating, Agentic AI focuses on doing
-
use cases
-
analyze market trends, assess investment opportunities, and even create personalized financial plans for individual clients.
- Optimizing logistics and supply chain management
Agentic AI
[Hugging Face definition] AI Agents are programs
where LLM outputs control the flow
(Tasks, their inputs, and sequence)
Characteristics
LLMs + Tools + Autonomy
- Multiple LLM calls
- LLM + Tools
- LLMs interact with each other
- LLM as an Orchestrator (above definition)
- Autonomy (LLM has autonomy as its generating unseen responses)
Anthropic
Agentic Systems
Agents are systems where LLMs dynamically direct their own processes and tool usage, maintaining control on how they accomplish their tasks
- Open Ended
- Feedback loops
- No fixed path
Risks:
- Unpredictable path
- Unpredictable output
- Unpredictable cost
Mitigation:
- Observability
- Guardrails are required
-
Probabilistic technology
Autonomous AI Agents
-
High adaptability to changing environments and events.
It relies on patterns and likelihoods to make decisions and take actions
-