Published: August 2025
One Last Note from Paper Pilot
Hey — I’m Rafi. This is the note I hoped I’d never write.
Paper Pilot is shutting down.
Here’s the thing.
January 2024, I landed in the U.S. with a backpack, a student visa, and a brain that only feels alive when it’s building. The first weeks were quiet in a way that didn’t feel peaceful. Classes, a tiny room, student loans, basic meals. No startup energy. No people to jam with. Just me trying to keep my head above water.
I couldn’t sit still. I kept calling a friend back home (India), passing ideas like a shuttle — insurance, logistics, anything that might stick. Most paths died fast.
Then AR glasses for warehouses pulled me in. I started reading research papers to see what’s real and what’s hype. That’s when I hit the wall: papers are dense, Full of jargon, time is short, and my head was tired.
So I built the thing I needed, Nothing grand. Just “make papers easier to understand.” A small, scrappy tool. No pitch deck. No funding plan. Just utility. I asked a friend who knew research to help.
From March to August, I taught myself what I didn’t know
Retrieval-Augmented Generation (RAG), which simply means pulling the right pieces first and then letting the model explain them. I mapped what data I could access and what I couldn’t. Rate limits. Dead APIs. I looked at what existed: Unriddle, Elicit, Consensus, Jenni. Useful, but narrow.
I wanted a real workspace, not chat-with-one-PDF, but a board where you could throw in papers, notes, reports, and talk to all of it together. Keep context. Remember decisions. Move faster. Up to 50 docs at once. That became the heart of Paper Pilot.
I went hunting for momentum. A month in Boston. Dozens of events. Sleeping on a friend’s floor. No cofounder, but that didn’t stop the build. On August 15th 2024, I shipped the first real MVP — research boards with multi-document chat. I sent it to the Google Gemini hackathon. Didn’t win. Still proud.
Then something unexpected worked.
We posted on Reddit for backlinks and ended up with users. People signed up because they saw the same problem. We crossed 1,000 users. A builder joined on the tech side. A designer came in. I walked into TechCrunch not as a developer, but as a founder. For a minute, it felt like the flywheel might actually turn.
Then NotebookLM launched.
Google-backed. Instant reach. Distribution you can’t match with sweat alone. They didn’t have to fight for attention. They had it from day one. You know how that story goes.
I didn’t fold. I pushed harder. Posters at the library. Pitched anyone who’d listen. Hired a marketing person; she left in a month. Entered UTA MavPitch Competition — aimed for $15k, landed $2.5k. Joined a pre-accelerator, Expected mentornship but got webinars and one-minute pitches. Met a mentor; had a couple of calls; nothing really moved. I kept showing up anyway, because sometimes that’s the whole job.
By August 2025, the math was clear: without funding, a full team, or a sharp differentiator, we weren’t going to out-build or out-distribute the giants. The gap wasn’t shrinking, and attention flowed to products with infinite channels. Paper Pilot helped people — just not at a scale that moved those lines.
I was out of runway and, if I’m honest, out of steam — totally burnt; the kind of tired a weekend doesn’t fix. I’d been living in survival mode for a year and a half.
So I’m choosing to stop. Not because I don’t believe in what we made. People used Paper Pilot, got value, sent thoughtful feedback, and even argued with me about feature priorities (my favorite kind of email). We shipped something real. But there’s a point where grit becomes denial. I know my limits, and I hit them.
If you used Paper Pilot, thank you. If you sent a bug report at 2 a.m., double thank you. To the friend who took my calls at weird hours; to the friend who helped with research; to the friend in Boston who gave me a couch and a shot; to the teammate who shipped with me when it wasn’t glamorous; to the designer who gave the product a face; to the random X account that helped with X marketing; to the mentor who took a couple of calls; to my roommates who kept me fed and gave me a place to stay; and to everyone who told me the truth when I asked for it — you mattered.
What this really means is: the chapter ends here, not the story. I’m working on the next thing — beonai.xyz — a fresh problem, a cleaner swing. I’m not promising fireworks on day one. I’m promising the same thing I’ve always had: curiosity, taste, and stubborn follow-through.
If Paper Pilot helped you, I’m glad. If it didn’t, I hope what comes next will.
Thank you for giving a small, scrappy tool a real shot.
— Rafi
Note — We’re deleting your content from our production databases by September 1st, 2025. We’ll keep only legally required records. If you need resources to help with your research, we've compiled a list of helpful tools and services on our resources page.
