Update ARCHITECTURE.md

ARCHITECTURE.md

@@ -7,65 +7,93 @@ AISA models agentic AI systems as composed systems in which behavior
 emerges from the interaction between reasoning, execution,
 infrastructure, evaluation, and governance.
 
+![Agentic AI Systems Architecture (AISA)](https://cdn-uploads.huggingface.co/production/uploads/676bac31dd95830fd9adf3cf/N6mUf5D7FzV5PXOl2Bm3z.png)
+
 ---
 
-## Layered Model
+## Layer Responsibilities (Summary)
 
-AISA defines seven architectural layers, each responsible for
-a distinct system concern.
+### Governance, Ethics & Policy Layer
+- AI policies and transparency standards  
+- Fairness and bias mitigation  
+- Privacy and data protection  
+- Accountability frameworks and human-in-the-loop governance  
+- Regulatory compliance and ethical reflection  
 
 ---
 
-### 1. LLM Foundation Layer
-Provides language understanding and reasoning capabilities.
-This layer focuses on interpreting inputs and generating outputs,
-independent of execution or control logic.
+### Development & Deployment Layer
+- Version control of agents and artifacts  
+- Continuous integration and deployment  
+- Performance benchmarking and A/B testing  
+- Cost–latency trade-off management  
+- Security, access control, and lifecycle tracking  
 
 ---
 
-### 2. Tool & Environment Layer
-Defines how the agent interacts with external systems.
-It mediates actions through controlled interfaces, separating
-reasoning from execution.
+### Evaluation & Feedback Layer
+- Component-level evaluations  
+- Behavioral monitoring and quality metrics  
+- Error analysis and prioritization  
+- Human-in-the-loop evaluations  
+- Automated regression testing  
 
 ---
 
-### 3. Cognitive Agent Layer
-Handles goal-directed behavior, including planning,
-decision-making, and memory access.
-This layer represents the core agent logic.
+### Agentic Infrastructure Layer
+- Workflow orchestration and coordination  
+- Multi-agent communication patterns  
+- State coordination and observability  
+- Logging, monitoring, and dashboards  
+- Latency, cost optimization, and EvalOps pipelines  
 
 ---
 
-### 4. Agentic Infrastructure Layer
-Supports execution, orchestration, and coordination.
-It manages workflow state, failures, and scalability.
+### Cognitive Agent Layer
+- Task planning and decomposition  
+- Reflection loops and self-improvement  
+- Memory management  
+- Multi-turn reasoning and goal tracking  
+- Integration of external feedback  
 
 ---
 
-### 5. Evaluation & Feedback Layer
-Monitors and evaluates agent behavior over time.
-It enables comparison, analysis, and continuous assessment.
+### Tool & Environment Layer
+- Tool creation and structured syntax  
+- Code execution and sandboxing  
+- Safe function calling and MCP support  
+- Error handling and retries  
+- Permission, rate control, and structured I/O  
 
 ---
 
-### 6. Development & Deployment Layer
-Governs system evolution through versioning,
-experimentation, and deployment control.
+### LLM Foundation Layer
+- Tokenization and inference  
+- Prompt engineering and instruction tuning  
+- LLM APIs and adapters  
+- Context window optimization  
+- Fine-tuning, alignment, and safety grounding  
 
 ---
 
-### 7. Governance, Ethics & Policy Layer
-Defines system-wide constraints, oversight,
-and accountability mechanisms.
+## Architectural Principles
 
----
+![AISA Architectural Principles](https://cdn-uploads.huggingface.co/production/uploads/676bac31dd95830fd9adf3cf/NZACvevXzxQR2dye4PNh_.png)
 
-## Architectural Principles
+**1. Separation of Concerns**  
+Clear separation between system responsibilities to prevent mixing reasoning, permissions, and orchestration logic.
+
+**2. Assurance-by-Design**  
+Evaluation, monitoring, and governance are built into the architecture from the start, not added after deployment.
+
+**3. Dual-Plane Design**  
+A clear distinction between the data plane (runtime execution) and the control plane (policies, permissions, and budgets).
+
+**4. Contract-Driven Interfaces**  
+Structured and machine-checkable interfaces reduce ambiguity and improve testing and auditability.
 
-- **Separation of Concerns** across system layers  
-- **Explicit Boundaries** between responsibilities  
-- **System-Level Evaluation** of agent behavior  
-- **Governance by Design**  
-- **Implementation Independence**
+**5. Continuous Improvement Loop**  
+Agent behavior evolves continuously through feedback-driven updates to prompts, tools, and policies.
 
+**6. Practical Deployability**  
+The architecture accounts for real-world constraints such as cost, latency, observability, access control, and versioning.