The leader track 🤝 · for partners, practice leads & independents · 6 x 45 min · no tools required
The jagged frontier
758 consultants, one study, two numbers pointing opposite directions: +40% quality inside AI's capability, -19 points outside it. What that does to the leverage pyramid, and why the bottom half gaining most changes who differentiates.
💷 Leader 2The engagement, re-priced
When drafts are nearly free, effort-based pricing leaks. "You used AI - why am I paying for three weeks?" answered honestly: value pricing, disclosure, and what PwC reselling ChatGPT signals about where fees go.
🛤 Leader 3The two rails
Samsung's leak and Deloitte Australia's refund, turned into policy: consumer vs enterprise tools as contract architecture, anonymization as habit, and a human owner for every fact that reaches a client.
🔍 Leader 4Quality at AI speed
Cheap drafts move the bottleneck to review. The partner as editor-in-chief: what to check when everything reads fluently, the review pyramid, and the anti-slop bar for work carrying the firm's name.
🏛 Leader 5The firm's AI operating model
Build, buy, or wrap: McKinsey's Lilli as the reference case, BCG's internal tools as the signal, and why the firm's proprietary knowledge - not its prompts - is the moat AI actually sharpens.
🧭 Leader 6Leading the transition
If AI does the analyst work juniors learned on, how does anyone become senior? The apprenticeship problem, the new bench curriculum, and the client conversation about AI use - held before they ask.
The consultant track 🛠 · for practitioners · 10 x 45 min · one engagement end to end · bring a real summary
Meet the engagement
The frontier numbers, the two rails, and the brief: Northbridge Advisory has six weeks to tell Cadence whether to launch an enterprise tier. Your governed setup and frontier map, built today.
📋 Consultant 2Proposal and scoping
The proposal AI drafts in minutes and the parts it cannot: scope judgment, the hypothesis tree, and pricing the work. Week 1 at Northbridge, with the workplan the next eight sessions execute.
🔎 Consultant 3Research that holds up
The invented statistic is consulting's fake legal case. Source logs, the lead-not-fact rule, and grounded research discipline - with the deep techniques pointed at learn-ai-research.
📊 Consultant 4Analysis co-pilot
Interview synthesis without flattening dissent, structuring data asks, and the findings-to-insights ladder - where AI accelerates and where it quietly averages away the interesting outlier.
🏛 Consultant 5The storyline
Minto's pyramid meets the live scorer: governing thought, action titles, and the ladder from the intern's 19 to a readout-grade 100 - one measurable move at a time.
🖼 Consultant 6The deck
One message per slide, exhibits that earn their place - and the anti-lever: "let AI write it all" produces beautiful prose that scores 8, because fluent is not specific. The measures cannot be charmed.
✉️ Consultant 7Client communications
Status notes, steering pre-reads, and the difficult-news email - drafted fast, owned personally. Voice is a client-anchor problem: AI defaults to everyone's voice, which is no one's.
🔒 Consultant 8The confidentiality drill
A draft that would score in the 90s - capped at 15 by the gate, because it names names. Anonymization patterns, enterprise-tier architecture, and the reflexes that keep your firm out of the news.
✅ Consultant 9QA and the readout
The citation audit (every fact, a human owner), the fact-check pass Deloitte Australia now wishes it ran, Q&A rehearsal with AI as the hostile client - and the readout itself.
🏁 Consultant 10 · capstoneCapstone: the engagement, compressed
The whole arc in one sitting, the ladder replayed from memory - then the paste box: your own executive summary, scored by the same measures, coached by the same table. Ship it.
The knowledge map 🧠
The whole course at a glance - hover a session to spotlight its concepts, click any node to jump in.