Skip to content

THE 37 QUESTIONS EVERY AI ARCHITECT SHOULD BE ABLE TO ANSWER

One diagnostic question per discipline, matching the "One Question" on each field card. If you can answer all 37 with specifics — not generalities — you think like an AI architect. Where you can't, that's where to look next.

Foundations

  1. Operating Model — Is this a decision I need to own, or one I need to enable someone else to make well?

  2. Model Ecosystem — If my primary model doubled in price or got deprecated next month, how much would it cost me to switch?

  3. Prompt Engineering — Can I roll back a prompt change in under five minutes, and do I have a test that would catch it if it broke something?

Knowledge Architecture

  1. RAG — When the system is wrong, do I check whether the right content was retrieved before I touch the prompt?

  2. Data Architecture — Is the data this AI needs actually ready — profiled, governed, and integrated — or am I assuming it is?

Agentic Systems

  1. Agents — What is the worst thing this agent can do before a human sees it, and is that limit in code or just in the prompt?

  2. Multi-Agent / MCP / A2A — Could one well-designed agent do this, and if not, what specifically requires the second one?

  3. Copilot Ecosystem — Before we turn this on for everyone, do we actually know what each user's Copilot can reach, and is that what we intend?

Security & Governance

  1. AI Security — If a user's input were treated by the model as an instruction, what's the worst it could trigger, and what limits that in code?

  2. Shadow AI — Is the sanctioned, governed way to use AI easier than the shadow way, or am I pushing people to work around me?

  3. Compliance & Model Risk — Could I show an examiner the inventory, the validation, the monitoring, and the owner for this model today?

Production & Platforms

  1. Observability & Evals — What's my faithfulness number, and would my eval suite catch it if this change made the system worse?

  2. Cost Engineering — What's the most this system could cost if every safeguard failed, and is that bounded in code?

  3. Infrastructure — At my actual volume and data sensitivity, does the break-even math favor API or self-hosting, or am I deciding by habit?

  4. Coding Assistants — Does AI-generated code get reviewed at least as carefully as human-written code, or are we trusting it more because it looks confident?

  5. Integration & AI-Native Design — Am I adding AI to this workflow or rebuilding the workflow around AI, and does my architecture match that answer?

  6. Platform Engineering — Do product teams use the platform because it's easier than building their own, or only because they're told to?

Landscape, Value & Direction

  1. Startup Landscape — For what I'm about to build, what's the moat, and could a funded startup out-build me and sell it to everyone?

  2. Business Value — What's the cost baseline, the target metric, and the name of the person accountable for moving it?

  3. Hype vs. Reality — Is this capability reliable in production conditions like mine, or have I only seen it be capable in a demo?

  4. Emerging Patterns — When the model stops being a differentiator in 18 months, what is my organization's actual AI advantage?

The Practical Layer

  1. The Practicum — Could I sit down right now and actually build this, or do I only understand it well enough to talk about it?

  2. Practice of Architecture — Does my design review actually change designs, or does everyone nod and ship what they already built?

  3. Production Operations — If the AI started giving wrong answers right now, how would I know, and how fast could I roll back?

  4. Commercial & Legal — Does this contract say our data won't train their models, and can we leave with our data in 12 months?

  5. Requirements Engineering — Does this requirement say what "good enough" means as a number, and what happens when the AI can't meet it?

  6. Organizational Patterns — Who loses their bonus if this AI initiative doesn't move its business metric — and if the answer is "nobody," why are we starting?

  7. Technical Debt — Which of my AI systems would fail an audit, a model deprecation, or a quality review today, and am I tracking that as risk?

  8. Communication — Does my diagram show what happens when the AI fails and where the PII goes, or just the happy path through a magic box?

  9. Career & Development — In the hard moment — deadline pressure, an over-promising vendor, a risky shortcut — am I the architect who holds the standard or the one who caves?

Cloud & Production Reality

  1. Cloud Platforms — If I had to migrate providers in 12 months, would my agent logic move with a config change, or need a rewrite?

  2. Agent Implementation — Is this agent's scope narrow enough to debug, its context lean enough to reason, and its dangerous actions blocked in code — or am I shipping the kind that fails?

Responsible AI

  1. Responsible AI — Who could be harmed by this system working exactly as designed, and does that person have a voice in the decision about whether to build it?

Advanced Topics

  1. Fine-Tuning — Is this gap about what the model knows or how it behaves, and do I have the baseline that would prove fine-tuning closed it?

  2. Multimodal — What information in our inputs lives outside the text, and how would I know if the system were silently getting it wrong?

  3. GraphRAG — Which real, recurring user questions fail today because the answer lives in relationships across documents, and who will own the graph that answers them?

  4. Model Portability — If my largest model provider gave me ten days' notice tomorrow, which capabilities would move by configuration, and which would become a project?

If you answered all 37 with specifics, you don't just know AI architecture — you practice it. Each question maps to a full module in the AI Architect's Comprehensive Reference.