About the guide

Agentic Architecture Skills is a practical guide to having AI agents do company work safely and usefully. It connects business opportunities with architecture, useful knowledge, operating practices and evidence. It rests on three ideas: the layers you own, the lifecycle every agent runs through and the capabilities you add.

About the author

Murali Sid works across AI, enterprise architecture, integration and product development. His experience includes electricity distribution, water utilities, energy, mining safety, compliance and commercial solution delivery. Those settings inform the examples, alongside broader enterprise and software-product needs.

Connect with Murali on LinkedIn.

Choose your starting point

• Start here for the whole route, in seven steps.
• Use cases for business opportunities and practical starting scopes.
• Architecture for the layers, the lifecycle and the capabilities around an agent.
• Intelligence ladder for choosing how to improve an agent's capability.
• Knowledge and memory for representing and using evidence.
• Research for findings, limitations and open questions.
• Skills for applying the guide to a real task.

How the guidance is maintained

Existing material is a starting point for review, not an unquestionable authority. Recommendations change when stronger evidence or a simpler approach warrants it. External citations identify the underlying source. Original experiments are attributed to the author, and proposed applications are labelled as proposals.

The examples do not imply that every proposed agent has been deployed. Product claims and vendor-reported outcomes are kept apart from independent evidence and local measurements.

Reuse

Content is licensed under CC BY-SA 4.0. Attribute reused material to Murali Sid and link to the relevant public page. Software is licensed under MIT.