Veirox
Request access
Self-hosted · sold, not self-serve

Runs on your infrastructure.
Never on ours.

There's no Veirox-operated cloud to sign up for. You install the full product — Docker Compose or Kubernetes — on infrastructure you already control, and trial it against your own data before you ever pay for it.

Request access See deployment options

How access works

Three steps, no rate card.

1

Request access

Tell us about your infrastructure and use case below. We hand you private, credentialed pull access to the install artifacts — a Docker Compose bundle or a Helm chart, your choice.

2

Install and trial

docker compose up or helm install on your own infrastructure. Full product, no feature gating, running against your own data from minute one. We're hands-on through the evaluation.

3

Decide, then license

Evaluate on your own timeline. When you're ready, we agree licensing terms directly — scoped to your team size and deployment, not a self-serve rate card.

Deployment

Bring your own infrastructure.

Both install paths ship the same product — API, workers, console, and (optionally) a bundled Postgres and Redis. Pick whichever matches how you already run things.

Docker Compose bundle

A single docker-compose.yml — API, worker, beat scheduler, console, and a small nginx proxy fronting everything on one origin. Add your own Postgres and Redis, or run the bundled ones. Best for a single-host evaluation or a small production footprint.

  • Guided install.sh — validates your .env, pulls images, brings the stack up, waits for health
  • Post-install smoke test confirms every service, the migration, and a real login round-trip
  • Upgrade is docker compose pull && up -d — migrations run idempotently on every start

Kubernetes / Helm chart

A production-oriented Helm chart — Deployments for API/worker/beat/console, optional bundled Postgres and Redis as StatefulSets (or point at your own managed databases), RBAC, NetworkPolicy, and PodDisruptionBudgets. Best when you're already running a cluster.

  • Bring-your-own external Postgres/Redis via a Secret reference — no plaintext credentials in values.yaml, guarded at render time
  • A pre-install Job runs database migrations before the API Deployment rolls out
  • Namespace-scoped RBAC, a default-deny NetworkPolicy, and PDBs so a node drain doesn't take the whole instance down

Both bundles are distributed privately, per customer — request access for credentialed pull access, not a public download.

What ships

One instance. Nothing tier-gated.

There's no free tier to upsell from, so there's no reason to hold anything back. Every install gets the full product.

Omnichannel agent

Web chat, Slack, Telegram, WhatsApp, and voice — one agent, full conversation history, streamed responses.

Tasks & webhooks

Cron, manual, and webhook-triggered automation. 9 webhook providers with signature verification, routing rules, and alert correlation.

Human-in-the-loop approvals

Must-always / must-never rules, signed one-click approval links, full reasoning and evidence attached to every decision.

Users & RBAC

Invite your whole team, role-based access control, audit-log export, and MFA — included, not a paid add-on.

Governance

Per-entity retention policies, GDPR-style data-subject export/erasure, PII inventory and auto-redaction on webhook ingest.

Veirox Connect

A tiny outbound-only agent that tunnels the platform into network-isolated infrastructure — no inbound firewall holes required.

FAQ

Common questions.

A Docker host (or a Kubernetes cluster for the Helm path), an S3-compatible bucket for file storage (AWS S3, Cloudflare R2, or self-hosted MinIO), and your own Anthropic API key or OAuth credential — Veirox is bring-your-own-model-key, so there's no platform-level LLM fallback to configure around.
Access is credentialed per customer — the install bundles aren't on a public registry, so a short access-request conversation comes first. From approval to a running instance is typically minutes, not weeks.
No — there's no Veirox-operated hosting option. The instance runs entirely on infrastructure you control, which is also why your data never has to leave your network in the first place.
Self-hosted · private trial access

Tell us about your infrastructure.

A few details on your stack and use case — we'll follow up with credentialed access to the install artifacts.

Request access Read the deployment guide