The team AI operating snapshot

People — who can do the work

When someone spends a few hours learning a better AI workflow, how does that land?

It reads as slacking — delivery pressure winsFine in theory; in practice nobody has the timeIt's expected — learning time is part of the job

If your most AI-fluent person left tomorrow, what would happen?

Most of the know-how would leave with themSome habits would survive; the momentum wouldn'tThe work would carry on — the fluency is shared

How does someone new learn how your team uses AI?

Trial and error, on their ownThey shadow whoever seems to knowThere's a shared, current starting point

Process — how work moves

Where does AI actually sit in how work gets done?

It's individual — each person uses it their own wayA few workflows depend on it, informallyIt's part of named workflows with clear handoffs

When AI-assisted work passes between people, does the context travel with it?

No — context gets rebuilt at every handoffSome of it, scattered across chats and docsYes — handoffs carry the context by design

Could you say what AI changed about your team's output last quarter?

Honestly, noWe have anecdotes, not measuresYes — specific, measured changes

Tooling — what the work runs on

When AI-generated docs or notes conflict, is there one place the team trusts as current?

No — versions pile up across chats and filesSort of — someone burns hours keeping things in syncYes — one source of truth, and AI works from it

Where does AI-generated work end up?

In personal chats and tabs — it evaporatesPasted into docs, when someone remembersIn the systems where work already lives

Do your AI tools see the context the work needs — docs, code, decisions?

No — everyone starts from a blank promptSometimes, when people paste it in by handYes — shared context is connected and reused

Governance — what keeps it safe

What are the rules for what AI can and can't touch — customer data, code, contracts?

There aren't any written downA policy exists; few could quote itClear, current, and actually followed

How does AI-assisted work get reviewed before it ships?

Same as always — we hope someone catches itExtra scrutiny, applied unevenlyReview points designed for AI-assisted work

If AI got something badly wrong last month, would you know?

Probably not, unless a customer told usWe'd hear about it eventually, informallyYes — the work is observable and owned

Where should we send your results?