The questions that keep coming back, the place to start and the habit that finally moves the bottleneck off your desk.
You’re the person everyone asks. The expense process, the cleaning contact, the rule for invoicing – none of it is really your job, but it all lives in your head, so it all lands on your desk.
The instinct is to build a wiki, but the reality is wikis don’t get read. Your team will still come to you, because asking is faster than searching. The fix is to make asking you not the fastest option.
Alexandra Buys is co-founder and COO of The Delta, a venture ecosystem that started in Cape Town and now runs a 12,000m² startup campus in Berlin. She scaled The Delta’s ops team from zero to 35+ and led the rollout of Claude across the entire business.
And she solved the human-FAQ problem for her finance manager first.
The move: Answer once, let AI deliver it forever
Stop being the answer. Be the person who decides what the answer is (once) and put it where AI can hand it to whoever asks next.
“Every time someone asks you something, quickly drop the answer into the context so next time they can just ask Claude.”
How to stop being your team’s human FAQ
1. Track which questions keep coming back
For one week, write down every question your team brings you. Not the big strategic ones; the small operational ones — process, contacts, rules, who-do-I-talk-to.
The repeats are the bottleneck. Each question you answer more than twice is a question that shouldn’t be coming to you at all.
2. Pick one area where the questions cluster
Don’t try to document everything at once. That’s the boil-the-ocean trap, and it’s why most knowledge bases die.
Pick the area where your week is bleeding most; usually, finance, ops, or whatever sits in the middle of your team’s daily work. Start there. Everything else can wait.
3. Write the answers in a place AI can actually use
Short, structured, one topic per file. Written for an AI to retrieve, not for a human to read end-to-end.
Your existing knowledge base (Notion, Confluence, Google Docs) isn’t the right surface for this. Those docs are written for humans, full of context AI doesn’t need, and the AI burns tokens hunting through them. A purpose-built layer of short markdown files works far better.
“Structuring things for AI is far more efficient.”
4. Make AI the first stop, you the second
The new rule: AI first, you second. When someone asks you something, your first response is to point them at the AI. If the AI doesn’t know, you answer once; and then add the answer to the context layer so the next person doesn’t need you.
Two things happen over time: The AI gets smarter at your business and the questions to you dry up. The bottleneck moves off your desk.
The big payoff
Your team gets faster answers than waiting for you. New hires onboard against the same context that the rest of the team uses. And the time you used to spend being the human FAQ comes back — for the work only you can do.
It takes a week of tracking and an afternoon of writing to start. It pays you back every week after that.
Want the full playbook?
This habit is one piece of Building Your Business’s AI Context Layer, Alex’s full masterclass inside the Founder Collab. The full session shows you the complete context-layer architecture The Delta runs on:
The full GitHub-based context layer structure (global folder plus per-team folders) and how it stays in live sync with Claude
The interview shortcut for building your first context doc from a one-hour recorded conversation
How to maintain the context layer through Claude Code with voice notes that auto-commit and broadcast changes to the team
The escalation rules pattern that tells AI when to handle something and when to loop a human in
The skills and plugins system that lets each team own its own AI tooling without breaking the rest
You’ll also get access to 40+ other masterclasses from SA founders and operators on sales, fundraising, UX, paid media and more inside The Founder Collab.
Get more SA tech and business news, tips and business-building workflows and subscribe to The Open Letter.
