AI and Solutions Architect Discovery Prep: Question Bank Design

The premise

SAs jump to architecture too fast; AI scaffolds a discovery bank that forces diagnosis first.

What AI does well here

• Draft tiered discovery questions by domain
• Suggest follow-ups that surface hidden constraints
• Format a discovery summary template

What AI cannot do

• Replace pattern recognition from real reps
• Predict the customer's political dynamics

Discovery is a question design problem

A weak discovery call is a demo with extra steps. A strong one ends with the architect understanding the customer's constraints better than the customer does. The difference is the question bank. AI can draft fifty discovery questions in seconds — the architect's job is to prune to the twelve that earn their slot, sequence them so each answer informs the next, and prepare the follow-up that turns a vague answer into a specific one.

A discovery anti-pattern checklist

• Asking questions whose answer you can find on the customer's public site
• Asking yes-or-no questions when an open-ended one would yield more
• Failing to ask about constraints — budget, timeline, internal politics, prior failed attempts
• Pitching the product before you understand the problem

A twelve-question bank that earns its slot

1. Walk us through the workflow this would change — what does it look like today?
2. Who else has tried to solve this internally? What did they build, and why is it still painful?
3. When you imagine this working in twelve months, what specifically is different — for whom?
4. What is the cost of doing nothing for another quarter?
5. What is the budget shape — capex, opex, or innovation fund — and who controls it?
6. Who has to say yes that we have not met yet?
7. What is the integration that, if it does not work, kills the deal?
8. What is your security review process, and what is its actual timeline?
9. What other vendors are you talking to, and what do you like about each?
10. If you had to pick our top concern about working with us, what would it be?
11. What does success look like in 90 days, and how would you measure it?
12. What would have to be true for you to sign by the end of this quarter?

Where AI helps and where it gets in the way

AI shines in the prep — drafting question banks, summarizing the customer's public footprint, sketching likely objections from competitors. It is dangerous in the call itself when architects start reading from a screen. Discovery is a relationship, not a script. Use AI to walk in with a sharper map; do not use it as a teleprompter. After the call, AI is useful again — clean transcript, structured notes, and a draft summary email back to the prospect within an hour. That follow-up speed often beats a polished proposal sent three days later.