Use this as a second-screen guide. The goal is to build trust, understand the business deeply, and leave with one clear next step, not to pitch every AI idea at once.
Listen first. Tony already asked for workflow insight. Your edge is not "Claude tips." It is mapping his real operation, spotting leverage, then showing where AI/product work fits.
Track authorizations, remaining units/hours, renewals, reassessment dates, payer rules, and schedule risk.
Flag missing signatures, vague notes, CPT mismatches, time mismatches, goal linkage gaps, and payer-specific requirements.
Structured parent/referrer intake, insurance uploads, eligibility checklist, missing-doc nudges, waitlist view.
Match client, RBT, BCBA, location, availability, authorized hours, and make-up sessions.
Onboarding, policies, payer quirks, documentation examples, supervision checklists, internal Q&A.
Intake, waitlist, staffing capacity, auth risk, cancellations, revenue leakage, and leadership visibility.
These are good if he asks, "What should I do with Claude right away?" Keep them as 5-minute education, not the whole meeting.
Do not give Tony the full Brandon links unless it feels relevant. Use them as proof of your approach.
Mostly listen today. Your ask should be permission to do a structured follow-up, not a big close.
"I usually invest a real discovery window up front. After that, I like to move into a small paid pilot so I can treat it like a real project and build something useful end to end."