Forge

The problem

Founders waste hours on the blank page. The business plan is the most-requested document of the early stage. Investors want it, accelerators want it, the founder’s own thinking wants it, and almost every founder I’ve watched stalls out on structure before they ever get to substance. Templates are either too long to be useful or too thin to be credible, and generic LLM prompts produce a generic LLM business plan.

The cost is not the writing. It’s the hours of blank-page anxiety a founder spends staring at section headings before the writing starts.

Before Forge

Founders started from a blank page or a borrowed template, spending hours on structure before they reached substance. Generic prompts produced generic plans, and no one challenged the weak claims before they were shown to an investor.

What I built

A business plan is a constrained document with a known shape. I use that shape as a schema and force the model to produce content that fits it, rather than asking it to write prose and hoping the structure emerges.

Then a second model challenges the first. One model writing in a single pass writes too confidently. A separate critic, with its own prompt and fresh context, catches the lazy paragraph, the unfounded claim, the executive summary disguised as a market analysis. One agent drafts, one critiques. The output is what survives both.

• Single-input intake. A one-line idea plus a few structured fields. No multi-step wizard.
• Streamed generation. The generator produces each section against a typed schema; sections stream in so the founder sees motion within a second.
• Critic agent. A second model call with its own prompt and context, emitting structured per-section feedback the founder can accept, edit, or ignore.
• Word & PDF export. The schema renders to a downloadable Word doc and a print-ready PDF. No accounts, no email gate.
• No persistence by default. The plan lives in the session. Nothing stored means nothing to leak, a feature, not an oversight.

The outcome

Live as a personal product, open to use, used by myself and a handful of founders for first-draft business plans. The most common feedback: the critic surfaced something they hadn’t noticed in their own thinking, which is exactly the bar I was building for.

Stack

Next improvements

• Template upload. Let founders drop in their accelerator’s house template so the schema renders to their exact headings.
• Persistence opt-in. Optional accounts for founders who want to iterate on a plan across sessions.
• Multi-format export. Add Markdown and Google Docs output alongside Word and PDF.