One screen.
One voice.
The whole show.

ShowPilot is an AI production copilot for live events. It replaces the clipboard, six browser tabs, and radio chatter with one dashboard, a voice agent (Hey Show) you can ask about the show out loud, and Org Brain — a portfolio-wide memory that answers questions across every tour, festival, and venue you run. You can also text it: Hey Show answers over SMS and WhatsApp with no app to install. It works for tours, festivals, and venues.

Hey Show is a voice agent built on OpenAI's Realtime API, running in the browser on your event dashboard. You press the mic — or the space bar — and ask about run-of-show variance, open incidents, weather and wet-bulb, crew status, SignalGuard threat posture, decisions, and tour memory. It calls a tool before it answers, never fabricates, and speaks the numbers back plainly. Every session is transcribed and saved to the event's comms log.

No. Anyone on the loop can text ShowPilot's number from any phone and get an answer from the relevant show's brain — over SMS or WhatsApp, no install, no login, no seat. The thread is sticky per number: it stays on your most recent show until you text SWITCH, so follow-ups like 'and the load-in time' just work. Text LIST to see your shows, SKILLS to run a saved macro by name or number, or GLOBAL to ask across the whole portfolio. Only verified numbers are answered — an unknown number is told to link itself in Settings first, and never gets show-specific data.

Yes — that is Org Brain, at /brain. It indexes every event you can see into one queryable graph: learnings, decisions, patterns, documents, linked feeds. Ask it in text or by voice, or text GLOBAL from your phone. Retrieval is hybrid keyword and vector with an LLM rerank and recency decay, and every answer carries a confidence band and cites the show it came from. When the portfolio does not cover the question, it abstains and tells you to widen scope rather than guess. It can also count: aggregate questions like 'which vendors run late most often' or 'how many weather holds last summer' return real grouped answers.

Yes, without anyone curating it. When a show closes, ShowPilot extracts the night's learnings automatically and runs cross-event pattern detection across that venue and that tour. The next upcoming show at that venue or on that tour opens with a pre-show playbook — the top five patterns by evidence count, ready before the crew briefing. Patterns that recur across many detection runs carry a stability signal, so the ones that keep proving out rise and the noise falls off.

Yes. ShowPilot ships an MCP connector, so AI clients can call its tools directly without the UI — list events, pull event detail, read tour memory and decisions, and read SignalGuard threat posture. The admin panel hands you a copy-paste config for Claude Desktop, Claude Code, ChatGPT, Codex, and the Anthropic API. Access is a per-org bearer token, hashed at rest, scoped to exactly the shows that org can see, and every call is audited. A cross-org request returns a not-found, never a permission error.

Yes. ShowPilot ships live integrations with Lennd (advance feeds + CSV uploads), OnTrack (incident feeds + per-module CSVs), Airtable, Google Sheets, Google Drive (with OCR for PDFs and images), SignalGuard threat-posture monitoring, and Weather / WBGT. Master Tour is currently advance-only with a live read coming.

Yes. The dashboard adapts to the integrations an event has connected, drawing from a catalog of 18 panel types — run-of-show, incident feed, weather and wet-bulb, crew, decisions, tour memory, SignalGuard, Lennd, OnTrack, Airtable, Sheets, Drive, and more. Build a layout by hand in the visual builder or describe it in plain English and have it generated. Save any layout as a template and it clones into every new event across your org tree, parent to company to venue.

All three. Per-Tour pricing covers theater, arena, and stadium tours. Per-Festival pricing covers weekend operations at events like Coachella, EDC, and Bonnaroo. Per-Venue pricing covers arenas, stadiums, and amphitheaters running their own calendar. There's also a Single-Show Trial ($5K) for one stadium date and a Portfolio License for Live Nation, AEG, and tier-one operators.

Single-Show Trial: $5K per show. Per-Tour: $50K (theater) / $90K (arena) / $150K (stadium). Per-Festival: $25K–$75K per weekend. Per-Venue: $2K–$5K per month per building. Portfolio License: $500K–$2M per year. All tiers include the full feature set; the difference is scope.

72 hours. You share a feed URL or grant access to your advance docs; we wire it up and validate against a live event before doors. Pilots typically run 30 days on one tour, one venue, or one festival weekend — no procurement, no PO, no six-month rollout.

ShowPilot is multi-tenant by construction. Every read is scoped to your org tree (parent, company, venue) server-side, and a cross-tenant probe returns a not-found, not a permission error, so one operator can never confirm another's shows even exist. Every comms exchange — voice, SMS, and WhatsApp — is transcribed, persisted, and exportable. Every decision is timestamped and searchable. Admin actions and super-admin impersonation are written to an audit log. The MCP API uses per-org bearer tokens, hashed at rest, with every request audited and exportable — the procurement-gate items legal and insurance ask for after a post-mortem.