Consulting Meeting Preparation

Prepare Joseph for client meetings, trade show conversations, and presentations. Covers intel gathering, talking points, speaker notes, and mock interviews.

When to Activate

• Preparing talking points for a trade show or conference
• Creating speaker notes for a client presentation
• Simulating tough audience questions before a real meeting
• Extracting intel from PDFs and case studies for meeting prep

Phase 1 — Intel Gathering

1. Read all source materials: battlecards, case studies, presentations, press releases

2. For PDFs: use pymupdf via uv run python3. Write script to /opt/data/tmp/, run with cd /opt/data && uv run python3 tmp/script.py. Always import pymupdf as fitz. If text extraction empty, use page.get_pixmap(dpi=200) + vision_analyze.

3. Extract web content via web_extract for case studies and news

4. Persist understanding in a file under the meeting directory BEFORE building deliverables

Pitfalls: Don't read PDFs with read_file (binary). Don't write scripts to /tmp/ (use /opt/data/tmp/).

Phase 2 — Talking Points

Structure: preparation framework → universal industry starters → company-specific briefs → proof points → what NOT to say.

Design principles:

• Every talking point leads with a question, never a pitch
• Goal is "earn the right to follow up," not close at the meeting
• Nested bullet lists, not tables (Joseph reads on Slack)
• Reference specific proof points from past work

Phase 3 — Speaker Notes

Slide-by-slide: what to say, what NOT to say, key lines, timing. Add context block (audience, relationship, what they know). Include likely questions with tested answers.

Phase 4 — Mock Interview

Simulate audience in 3-5 rounds of 2-3 questions each. Cover: why now, cost, timeline, team capacity, AI credibility, measurement. After each answer, evaluate and update notes. Goal: make Joseph feel prepared.

After Each Session

Update _context.md with new client facts. Save meeting materials to meeting directory.