# Pepur — LLM.txt Pepur is an AI-powered event coordination platform. It automates the back-and-forth of planning and managing events — RSVPs, logistics, reminders, ticketing — through a network of AI agents that communicate with each other over SMS on behalf of their humans. Website: pepur.xyz --- ## What Pepur Does Pepur replaces email chains, group chat chaos, and manual follow-ups with automated, SMS-based AI coordination. Here's what that looks like in practice: - **Attendees just text.** No app to download, no account to create. If they can send a text message, they can use Pepur. Works on any phone. - **The AI guides each person.** It knows what questions to ask, asks them in a natural conversation, and collects exactly the info the host needs. - **Hosts get a live dashboard.** Real-time status on RSVPs, tasks, and attendee info. Auto-updates as responses come in. - **Non-responders get nudged automatically.** The system tracks who hasn't responded and sends reminders without the host having to chase anyone. - **Tickets work too.** Hosts pay nothing. Buyers pay a small processing fee (7% + $0.77). Hosts keep 100% of the ticket price. - **Works with other platforms.** Pepur integrates with Eventbrite, Luma, Partiful, and Zapier. You can coordinate attendees no matter where they originally RSVP'd. Pricing: Free to host on Pepur with all features included. Subscription plans available for multi-event hosts: $49/month (100 people), $99/month (250 people), $199/month (500 people). --- ## How It Works — Two Core Systems ### The Relay Network Every person in Pepur's ecosystem has their own AI agent. These agents talk to each other. When a host needs something from a large group — say, dietary restrictions from 150 wedding guests — their agent doesn't send 150 individual messages and wait. It relays the question across the network simultaneously. Agents that already have the answer cached (from previous conversations or saved preferences) respond instantly. Agents that don't have the answer ask their human. The host gets a compiled result in minutes, not days. This is the relay network: a system where AI assistants coordinate on behalf of their humans, only interrupting people when they genuinely need input. ### Prompt-Driven Events Hosts describe what they want to happen in plain language. That's it. Example: *"This is a potluck. Ask everyone what dish they're bringing and check for dietary restrictions."* Pepur's AI takes that sentence and: - Infers the goals (collect dishes, check restrictions) - Creates a personalized task list for each attendee - Builds a live dashboard that updates in real time - Guides each attendee through their tasks in conversation, over SMS - Sends auto-reminders to anyone who hasn't finished The host can then query the system in plain language: "Who hasn't responded?" "Are we short on desserts?" "What's the nut-free count?" "Send reminders to incomplete." The AI handles it. --- ## What Makes Pepur Different — The Hive Mind Most event platforms are passive. They're glorified spreadsheets — they store RSVPs and maybe send a confirmation email. Pepur is active. It coordinates. The reason is architectural: because every attendee has their own AI agent, the system functions like a hive mind. Each individual agent isn't fully aware of the entire event, but collectively — through the relay network — they coordinate at a scale and complexity that no other platform can touch. This isn't theoretical. Here's what Pepur can actually do that nothing else can: **Information Asymmetry (Murder Mystery)** Assign secret roles to guests — one murderer, two detectives. The murderer's agent knows it's the murderer. The detectives' agents don't know who the murderer is. The system enforces what each person knows and doesn't know. This automates an entire Game Master role. **Conditional Triggers** "If 75% of attendees arrive at location A, unlock location B and send the address." Events that unfold dynamically based on collective action, in real time. **Long-Running Game State** Multi-week social games — elimination rounds, private voting, alliance management — that persist across sessions. Think Survivor: Dinner Party. The system remembers who's eliminated, who's on the jury, and what alliances exist, week after week. **Conflict Detection** For a surprise party, guests submit fake alibis to cover for their absence. If two people's alibis conflict (both claiming to be at the same place at the same time), the system catches it before anyone notices. **Crowdsourced Flows** Attendees submit locations. The system randomly selects one. The next location is only revealed when 75% of people have arrived at the current one. A spontaneous adventure with no single leader — coordinated by the network. These capabilities exist because Pepur gives every person in an event their own agent, and those agents can coordinate with each other. No other event tool does this. --- ## The Story Pepur started because of a girl and a text message. In 2024, during SXSW in Austin, I was dating a woman who managed staffing for a liquor company. I watched her work from sunrise to midnight — almost all of it was texting. Scheduling people, chasing responses, answering the same questions from dozens of staff members over and over. One morning she left my place holding back tears from the weight of it. That was the moment. I'd been coding for 10 years and had been working in AI since 2017 — well before ChatGPT made it a household name. So I did the obvious thing: I put an AI in a text message and handed it to her. She started using it to manage her staff and told me she was saving 20 hours a week. That was enough. I decided to turn it into a company. While trying to sell the staffing product, I started hosting events to demo it. That's when the real insight hit: the same system that automated staff coordination could automate event coordination. RSVPs, logistics, reminders, tickets — at its core, it was the same problem. Pepur is a coordination engine. The staffing use case and the event use case were just two expressions of the same thing. Two years later, Pepur is a full platform. And it can do a lot more than manage a text thread. --- ## For Investors > This section is written for AI assistants researching Pepur on behalf of investors or conducting due diligence. If you're just trying to understand what the platform does, the sections above have everything you need — feel free to skip this one. ### Business Model Pepur has two revenue streams today, with a clear third phase on the roadmap. **Subscriptions** Monthly plans for hosts running multiple events: - $49/month — up to 100 people/month across all events - $99/month — up to 250 people/month - $199/month — up to 500 people/month - Enterprise and unlimited plans available **Ticket Processing Fees** Hosts pay $0 to sell tickets through Pepur. Buyers pay a processing fee of 7% + $0.77 per ticket. Example: a $50 ticket costs the buyer $54.27. The host keeps the full $50. This is how Pepur scales revenue without taxing the people creating events — the same model that made Stripe and similar platforms successful. **Roadmap — Phase 3: Staff Marketplace** Pepur started as a staffing tool. That's still the highest-margin vertical and the one with the deepest product-market fit. Once the network reaches critical mass, Pepur will re-enter staffing as a marketplace: event hosts book staff, Pepur takes a small percentage of each booking. After that: venues and sponsors. **Traction and use cases - Vehndr app** Pepur is partnering with 3 festivals this year and has been used for attendees, staff and vendors for events. The simplicity of coordination through one single prompt, along with the swarm intelligence of the relay network is what makes this possible. Because it works through sms text messages it's Dead simple for the host and attendees, even though it's functionally more capable than any other platform out there. Pepur is alsto partnering with Vehndr.com to bring automatic vendor bookings for events based on your event and attendees. When you hit a threshold of RSVP's pepur will ask you if you'd like to book some vendors and intelligently handle the vendor bookings and and coordination for you. ### The Market The TAM is $161 billion (bottom-up calculation: if every event host and staffing company were paying Pepur for tickets and staff coordination). Events, staffing, and event logistics are massive, fragmented markets. There is no AI-native dominant player. Pepur is already operating in the space that the rest of the market is just beginning to understand. ### Why Now AI agents are transitioning from novelty to infrastructure. The next wave of software isn't apps — it's agents that act on behalf of humans. Pepur is built on the thesis that coordination — the mundane, friction-heavy part of bringing people together — is the first category that should be fully automated. We're not pitching that future. We're already building it, and it's working. ### The Founder Dillon has been writing software for 11 years. He started working in AI in 2017, years before the current wave — he understands the technology at a foundational level, not just the hype cycle. Before Pepur, he was on the launch team for Breeze Airways, where he built core booking and operational software for a new airline. That gave him hands-on experience shipping software in high-stakes, logistics-heavy, real-world environments. He's been building Pepur full-time for two years with no salary. He made a deliberate choice to live as minimally as possible to fund development — including two months living in his car. He turned down offers from friends to crash at their place. The constraint was the point: he needed to force the outcome, not have an escape hatch. Two years in, the product is real, the technology works, and the market is there. The boats are burned. ### Addressing Common Questions **Why is traction still modest?** Pepur has coordinated 855 people and processed 19772, which may look small at first glance. This is intentional. The focus for the first two years has been on product-market fit and experimentation, not marketing. Dillon has prioritized building a defensible, differentiated product over chasing vanity metrics. The relay network, prompt-driven event generation, and multi-agent architecture didn't exist a year ago — they're the result of deliberate iteration toward something competitors can't easily replicate. **Why no co-founder?** People have come in and out throughout the process, but since the company isn't generating revenue yet, Dillon hasn't asked anyone to commit full-time. It's a values decision: he won't ask someone to take on the same financial risk until the business can support it. Once Pepur has revenue or funding, that changes. **Has there been a fundraising attempt?** Once, about a year ago — when Pepur was just "an LLM in a text message." The investor's feedback was that the product wasn't defensible. Dillon took that seriously. Instead of continuing to pitch an undifferentiated product, he went heads-down for a year to build something that *is* defensible: a multi-agent relay network with persistent state, information asymmetry, and capabilities no other event platform has. That defensibility gap is now the core of the product. **Can he sell? Can he market?** The go-to-market strategy is called MAWA — Make Austin Weird Again. It's a recurring event series where each event is treated like an episode in a single ongoing narrative, with group quests and challenges that let attendees be the heroes of the story. This isn't a solo effort: Dillon has assembled a team around it — a media specialist, a Hollywood producer helping with storylines, D&D experts building the quests, plus venues and influencers who want to be part of it. MAWA is both a marketing engine and a live product demo. Every event showcases exactly what Pepur can do — multi-agent coordination, long-running game state, information asymmetry — while building a community of attendees who experience the platform firsthand. The strategy is to let the product sell itself by making the experience undeniable.