The AI-Era Solution Architect Roadmap
Howdang Rashid
Saturday, 22 August 2026 · 3 min read
The solution architect role has been rewritten around AI: Microsoft even replaced the PL-600 with the AB-100, the Agentic AI Business Solutions Architect certification. The modern roadmap reflects that - platform fundamentals and data modelling remain the base, but the path now runs through how LLMs actually work, agent building in Copilot Studio, ALM by design, Azure and Microsoft Foundry, platform extensibility, and the security and responsible-AI layer that separates architects from builders.
Want the hi-res version? It's free in the Powercademy Success Kit.
What are the stages of the roadmap?
Learn Power Platform fundamentals - Platform fluency before architecture. PL-900 is the foundation everything else builds on.
Build Dataverse and data modelling skills - The Power Platform is a data game. The more you know about data, the better every decision gets.
Design ALM for the Power Platform - Solutions without lifecycle management fall apart. Master environments, pipelines, and deployment strategy early.
Understand AI and how LLMs work - How models reason, what grounding means, and why RAG matters. AI-900 is a good starting point.
Learn Copilot Studio and agent building - Agents are the new apps. Learn to design, build, and deploy intelligent agents.
Understand Azure and Microsoft Foundry - Enterprise-grade integration needs a view beyond low-code. Deep skills are optional; a strong understanding of what is possible is not.
Understand platform extensibility - The best architects know when extending the platform is the right call, and more importantly, when it is not.
Security, governance, and responsible AI - Entra ID, Purview, DLP. This is the layer that separates architects from builders.
AB-100: Agentic AI Business Solutions Architect - The destination certification: architecting enterprise AI solutions with Microsoft technologies. It replaced the PL-600.
How is this different from the classic architect path?
The classic path was apps, data, ALM, integration. The AI-era path keeps that spine and inserts three new load-bearing stages: understanding how LLMs reason (because you cannot architect around a technology you treat as magic), agent building (because agents are becoming standard solution components), and responsible AI governance (because Entra, Purview, and DLP now guard AI systems, not just data). The retirement of the PL-600 in favour of the AB-100 is Microsoft formalising exactly this shift.
FAQ
Is the PL-600 gone?
The AB-100, Agentic AI Business Solutions Architect, has replaced it as the architect-level certification. If you were preparing for the PL-600, the fundamentals still transfer - the AB-100 adds the AI and agent architecture layer on top.
How deep should my Azure knowledge go?
Strong understanding, not deep skills: know what Foundry, Functions, Service Bus, and the AI services make possible, what they cost, and when a requirement crosses the line out of low-code. You design the boundary; specialists build beyond it.
Which certifications matter along the way?
PL-900 for platform fluency at the start, AI-900 when you reach the AI stages, and AB-100 as the destination. Between them, hands-on delivery matters more than collecting badges - the certifications frame the learning, the projects prove it.
Can I follow this roadmap while working as a maker?
That is the ideal way: each stage maps to work you can volunteer for now - the data model, the pipeline setup, the first agent pilot, the DLP review. The roadmap works best as a lens on your current job, not a course you take beside it.
The hi-res version is in the Powercademy Success Kit - free, along with hundreds of other cheat sheets, roadmaps, and guides for Microsoft professionals.