The Human × AI playbook running alongside you while you do it.
Methodology-native. Persistent across decisions. Built around the four moves: capabilities, workflows, judgment, learning loops. In the room with you on the actual work, every week, every quarter.
The Human × AI playbook in conversation form, held against the work in front of you. Methodology-native. Persistent across decisions. Five named moves are how it behaves in a real working session.
The leader. I need to walk the board through our AI strategy before the raise. Honestly, I have the technology story. I don't have the company story yet.
The companion. Then we're not starting with the deck. We're starting with your customer. What will never change about what they want from you?
The leader. (pause) They want to learn faster. They want it easier. They always want it cheaper. That doesn't change. With or without AI.
The companion. Good. Now name what AI is actually changing. Not in your product. In your economics.
The leader. The cost of production. It's collapsing.
The companion. Then your strategy isn't an AI strategy. It's three redesigns the cost shift forces on you. One. What your product becomes when production is cheap. Two. How your teams deliver, capability by capability, with AI in the workflow, not bolted onto it. Three. The cost structure of the company itself, because it has to evolve when the cost of production does. Take them in that order.
The leader. (longer pause) That's actually what I've been working through underneath. I just haven't said it that way to the board.
The companion. Then say it that way. The board narrative isn't a reframe of the deck. It's the sequence you just named. The AI work is the proof you're executing it. Move the technology to the appendix. Put the three redesigns on the cover.
She walked in with an AI strategy. She walked out with three redesigns and the order they belong in. The deck reorganised itself after that.
Not faster at the work you already know. Different order on the decisions that matter.
The leader. My VP Sales is pushing for Germany next quarter. We've got the AI worker generating outbound there already. Three inbound leads from Munich last month. He wants to scale it.
The companion. The AI worker found you three Munich IPs. That's the tool working. It's not a buyer signal. Scale outbound now and you'll spend a quarter generating volume against a buyer you haven't named. That's the failure mode the AI accelerates, not solves.
Before you scale: name the German buyer. Title, company size, the budget line they'd pay you out of. Then re-aim the AI worker at that buyer specifically. Same tool, different instruction set. The market follows the buyer; the AI worker follows both.
One scale decision held. The AI worker re-aimed before it was scaled.
The leader. I'm replacing my VP RevOps. The recruiter sent me five profiles. All senior, all from comp companies. I'm interviewing two on Friday.
The companion. You're hiring the role you had, not the role the function needs. Half the work your last VP RevOps did is now AI-doable: pipeline hygiene, forecast roll-ups, territory coverage analysis, deal-stage flags. If you backfill at the same shape, you'll pay senior comp for work the system can already absorb, and the gap that's actually open won't get filled. That gap is building the learning loop between the AI worker and the team.
Before Friday, write the capability you actually need: what should this function be able to do with AI in the system that it couldn't do before? Then look at the five profiles against that. My guess is two of them drop off the list and a different profile entirely becomes the actual hire. Someone who has built revenue ops with AI, not on top of it.
Two interviews postponed. The capability written first. The shortlist redrawn.
The leader. We've moved LinkedIn outreach to autonomous. The team wants the time back for higher-judgment work.
The companion. Look at it twice. That outreach is one of the few moments where the leader's judgment is actually on the wire. What to say, to whom, with what specificity. Autonomous strips the judgment that was the reason it worked. Reclassify as augmented.
One workflow pulled back from autonomous. Judgment placed deliberately, not residually.
Three questions. Three reframes. One week.
The list doesn't get shorter. The order changes. The order is what compounds.
Karine designed the companion to do exactly this work, and stays close to the leaders inside the cohort.
Across the redesigns in Karine's track record at Chegg and her current advisory work in tech and tech-services, the pattern is consistent. The ranges below are what becomes possible when the work is governed before the AI is.
P&L cost reductions of 10–25%. FTE reductions of 15–30%, classified, not cut. Innovation delivery 5–10× faster. Customer validation pulled from months to weeks.
Directional ranges drawn from operator-grade engagements, not industry benchmarks.
Companies that run the Human × AI playbook don't deploy a tool. They redesign the work, and everything that follows changes with it. Margin expansion alongside value creation. Cost-out and growth in the same motion, not as a tradeoff.
If you're not ready to reserve a seat, start where everyone else did. Read the playbook as it gets written. One essay a week from inside a real transformation.