Why we built Ghostpen
The blank-page problem
Most AI writing tools hand you a blank textarea and ask you to be your own prompt engineer. That's the wrong product for someone who just wants to write a cold email to a recruiter or a cover note for the role they actually want.
I noticed this watching friends try to use ChatGPT, Jasper, Copy.ai. They'd open the tool, stare at the empty box, paste in some random context, get back generic copy that sounded like every other AI email — "I hope this message finds you well, I am passionate about your company's mission" — and close the tab. Worse: some shipped that copy and never heard back.
The problem isn't AI quality. It's that the products treat writing as a prompt-engineering challenge instead of an outcome-engineering one.
What we did differently
Ghostpen templates ask for structured inputs. The investor-cold-email template doesn't say "describe what you want." It asks for the investor's name, their fund, your company in one line, your specific traction, and exactly what you're asking them for.
The AI gets clean, structured signal. You get back something that reads like a founder who values everyone's time — because that's what the system prompt enforces, and that's what the structured form scaffolds you to deliver.
The bet is that 90% of high-stakes business writing fits this shape. Cold emails. Cover letters. Investor updates. Follow-ups. Resignation letters. Recruiter outreach. Each has a known structure that good operators learned through reps. The template encodes those reps.
What's next
We're shipping outcome-driven templates for the work people actually do — job applications, investor outreach, sales follow-ups, networking. One credit per generation. Credits never expire. Free tier: 25 credits, no time-bound.
If you've ever written 30 cover notes hoping one would land, or sent a cold email that took an hour to draft and went unread, we built this for you.