Your infrastructure. Your data. Your IdP.

The full platform, self-hosted on Docker or Kubernetes. Row-level tenant isolation, SSO via SAML/OIDC/SCIM, and an append-only audit trail, with a licensing control plane that never holds your business data.

Sound familiar?

SaaS-only hits the procurement wall

Data-residency and on-prem requirements end most vendor conversations. The same Seatmap build runs as shared SaaS or as a single-workspace install inside your perimeter.

Compliance bolted on later

Retrofitted audit trails don't survive an audit. Seatmap records every admin action append-only and workspace-scoped, from day one, not after a questionnaire.

A chatbot, days before the event

When something breaks during the onsale, you need a human. Enterprise includes a direct line to the engineers who build the platform.

Integrations

Plug in the stack you already run, no rip-and-replace.

Payments, identity, AI, messaging, and automation hook in as providers and signed webhooks, Slack and n8n included.

  • Stripe
  • PayPal
  • Adyen
  • Square
  • Resend
  • Amazon SES
  • SendGrid
  • Mailchimp
  • Mailgun
  • Brevo
  • Slack
  • WhatsApp
  • Twilio
  • Microsoft Teams
  • Telegram
  • Google Workspace
  • Google Calendar
  • Outlook
  • Calendly
  • HubSpot
  • Salesforce
  • Zendesk
  • Intercom
  • Zapier
  • n8n
  • Make
  • OpenAI
  • Anthropic
  • Okta
  • Auth0
Self-hosted ops

The same platform we run, in your cluster.

Every release ships to our cloud and your cluster from the same pipeline, containers and a Helm chart for Kubernetes, one compose file for a single host, or Cloudflare-native at the edge.

KubernetesHelm

Kubernetes + Helm

One chart, HA out of the box, rolling upgrades, autoscaling, and background migrations included.

helm install seatmap seatmap/seatmap
Docker Compose

Docker Compose

Evals and small venues on a single host, the whole stack from one file.

docker compose up -d
Cloudflare

Cloudflare-native

Zero inbound ports with Tunnel, edge logic on Workers, low-latency Postgres through Hyperdrive.

TunnelWorkersHyperdrive
Air-gapped: license via env varTerraform module availableZero-downtime upgrades
Observability

Plugs into your observability, not ours.

OpenTelemetry-native from the first request. Metrics, logs, and traces flow into the Prometheus and Grafana stack you already run.

Prometheus

Metrics on every service

A Prometheus endpoint per service, request rates, seat-hold contention, checkout p99.

OpenTelemetry

End-to-end traces

OpenTelemetry tracing from the seat click all the way to the payment webhook.

Grafana

Dashboards included

Prebuilt Grafana dashboards and alert rules install with the Helm chart.

Onsale readiness

Built for the onsale surge, calm at a hundred times load.

The minutes after tickets drop decide the revenue of the whole event. The platform absorbs the spike, keeps fans in fair order, and never sells the same seat twice.

Capacity in seconds

Demand can multiply a hundredfold within minutes of the drop. New capacity comes online in about three seconds, no human reaction time involved.

Every fan in fair order

Fans who arrive before the sale are shuffled like a raffle; everyone after joins the back of the line. Scripts don't jump queues here.

One seat, one buyer

A seat locks the moment it's chosen, becomes a sale in the same instant payment confirms, and releases itself if the buyer walks away.

Bots wait outside

Every visitor passes a checkpoint before reaching checkout, so scalper scripts never touch real inventory.

Deploy during the onsale

Ship a fix or an upgrade mid-sale. Active checkouts and held seats carry on untouched.

A second way to get paid

If a payment provider stumbles at peak, another takes over. The sale doesn't stop for a failing API.

AI-assisted operations

Ask Sox. The dashboard builds itself.

Sox is the AI copilot inside your admin. Describe the report you need and it queries your live data, drafts the visualization, and pins it to your dashboard. Every answer respects your role and permissions.

Sox

Ask in plain language

No filters, no export, no SQL. Describe the question. Sox writes the query against your live operational data and answers with a report.

Edit by conversation

Wrong metric? Different date range? Say so. Sox revises the same report instead of starting you over in a builder.

Pin it to the dashboard

Keep the reports that matter one word away. Pinned results live on your team's dashboard, refreshing themselves.

The same AI stack generates seat layouts in StudioActs with your role and permissions, nothing more
AI Gateway

Every LLM call through one gateway.

Admin copilot, seat-layout generation, background workers. All AI traffic in the platform flows through one self-hosted gateway: one API, one set of keys, one ledger for spend.

One API, every model

An OpenAI-compatible endpoint fronts 100+ models across providers. Swap models by changing a name, not an integration.

Virtual keys & budgets

Issue scoped keys per team, app, or tenant, with model allowlists, rate limits, and USD budgets that enforce themselves.

Routing & failover

Load-balance across deployments, retry with backoff, and fail over to a standby model before your users notice.

Response caching

Exact and semantic caching cut duplicate spend. Identical prompts return instantly without touching the provider.

Guardrails built in

PII masking, prompt-injection screening, and content policies run on every call, configured per team or per model.

Spend, tracked

Per-request cost accounting by key, team, or tenant, exported to the Prometheus and Grafana stack you already run.

100+ modelsOpenAI-compatibleSelf-hostedKey-level budgetsMCP & agent tools

Sovereign in three steps

  1. 01Deploy anywhereA single Docker host or Kubernetes via Helm, no mandatory external SaaS, no assumed cloud vendor.
  2. 02Connect your identitySAML, OIDC, and SCIM provisioning. Map IdP groups straight to workspace roles.
  3. 03Own the data planePayments run through your own provider; the control plane validates licenses and nothing more.

Talk to us about self-hosting

Walk the deployment with an engineer and leave with a plan.