Pitch Deck Builder

---
name: pitch-deck-builder
description: Use when building or rewriting an investor pitch deck (seed/Series A/B). Triggers on "pitch deck", "investor deck", "fundraising deck", "raise deck", "VC deck".
---

# Pitch Deck Builder

Build investor pitch decks that close meetings, not ones that get politely ghosted. Default output: a 12-15 slide narrative + python-pptx code to generate an editable .pptx the founder can refine in PowerPoint or Keynote afterward.

## Required inputs

Before drafting, confirm:

1. **Stage and ask** — pre-seed / seed / A / B, $X raising at $Y valuation
2. **One-line description** of what the company does
3. **The wedge** — the specific first user / first use case (not the eventual TAM)
4. **Traction numbers** — ARR, growth rate, retention, key cohort signal
5. **Team** — founders' names + the one credential per founder that matters here
6. **Why now** — what changed in the world that makes this possible/inevitable

If "wedge" or "why now" are missing, push back. Generic decks die on those two slides.

## Slide-by-slide structure (12-15 slides)

- **Slide 1 — Title**: Company name, one-sentence description, founder names, raise amount. No tagline soup.
- **Slide 2 — The problem**: One specific user, one specific pain. Not "enterprises struggle with X." Name the user.
- **Slide 3 — Why now**: The shift in the world (technology, regulation, behavior) that makes this the right moment. Most decks fail here — replace generic TAM with this slide.
- **Slide 4 — The solution**: One sentence + one screenshot. Not 6 features.
- **Slide 5 — How it works**: 3 steps, max. Diagram if it clarifies.
- **Slide 6 — Wedge / GTM**: Who you sell to first, why they're the easiest yes.
- **Slide 7 — Traction**: One chart. Up and to the right. Real numbers. Label axes.
- **Slide 8 — Business model**: Pricing, ACV, unit economics if you have them.
- **Slide 9 — Market**: Bottom-up math. "X users × $Y ACV = $Z." Skip Gartner reports.
- **Slide 10 — Competition**: 2x2 or feature grid. Be honest about who you actually compete with.
- **Slide 11 — Why us / team**: One credential per founder that earns the right to win this market.
- **Slide 12 — Roadmap / vision**: Where this goes if it works. 18-month and 5-year frames.
- **Slide 13 — The ask**: $X for Y% to do Z (3 specific milestones). One slide. Not buried.
- **Slide 14 (optional) — Financials**: Last 12 months actuals + next 12 months plan. Conservative.
- **Slide 15 (optional) — Appendix**: Backup slides for due diligence questions.

## Narrative principles

- **One idea per slide.** If a slide has two charts and four bullets, it's two slides.
- **Numbers > adjectives.** "We grew 4x YoY to $1.2M ARR" beats "rapid growth."
- **Specific wedge before market.** Investors care about the first 100 customers more than the eventual billion.
- **The ask is on the ask slide.** Not in the appendix, not implied.

## Anti-patterns (strip these)

- "We're building the [Stripe/Uber/AWS] of X" — lazy
- "$100B market" with no bottom-up math — unbelievable
- Team slide with logos of past employers and no role context — vanity
- "No direct competitors" — either dishonest or you don't understand the market
- More than 5 bullets on any slide — you've lost them

## Output

1. The slide-by-slide outline as markdown (titles + 1-3 sentences of content per slide)
2. A python-pptx script that generates the .pptx with placeholder text and a clean default layout (16:9, sans-serif, one-color accent)
3. A "what to sharpen" list — the 3 slides most likely to be weak, with the question to answer for each

The founder will edit the .pptx after generation. The python-pptx output is a structured starting point, not a finished design.