Enterprise

Your machines. Your people. Your rules.

Run the agent inside your own network. Let your team — and your clients — prompt it with roles. Keep every action on record.

SSO / SAML SOC 2 Type II VPC & on-prem 99.9% uptime SLA
Deployment

Your code never leaves.

The agent works on your servers, against your repos, behind your firewall. Orquesta coordinates; your machines do the work.

  • Managed cloud, or your own VPC on AWS, GCP or Azure
  • On-prem via Helm or Docker Compose
  • Air-gapped, with offline agent packages and your own models
Agents · your network3 online
build-box-01on-prem · Linuxrunning
staging-vpcAWS us-east-1idle
ci-runner-02GCP privateidle
Roles

Everyone asks. Not everyone ships.

Developers, PMs and your clients prompt the same agent. Roles decide what each can see, ask and deploy — per project and environment.

  • Eight built-in roles, plus custom ones
  • Command allowlists and blocklists
  • Two or more sign-offs before production
Who can do what
Org Ownereverythingfull
Developerprompt · dev + stagingship
Clientprompt · staging onlyneeds approval
Viewerread logsread-only
Live control

Watch it work. Take the wheel.

Every session streams live. When something looks wrong, a person takes control of the same terminal, on the same machine — then hands it back.

  • Mandatory diff review before execution
  • Auto-reject after a configurable timeout
  • Approval pings in Slack or Teams
Live session · build-box-01you're driving
$ npm run migrate -- --env prod ! 0042 would drop invoices.legacy_id agent paused — waiting for a human
Approved by 2 of 2 reviewers
Handing control back to the agent
Audit

Every prompt, every command, on the record.

Who asked, what ran, on which machine, and who approved it — in an append-only, cryptographically chained log you can stream to your SIEM.

Capabilities

Everything IT will ask for.

SSO & identity

SAML 2.0 and OIDC with Okta, Entra ID, Google Workspace. SCIM provisioning, IdP group-to-role mapping, enforced MFA.

Roles & permissions

Eight built-in roles from Org Owner to Viewer, custom roles, environment-scoped access and command allow/blocklists.

Flexible deployment

Managed cloud, your VPC on AWS, GCP or Azure, on-prem via Helm or Docker Compose, or air-gapped.

Immutable audit trail

Append-only, cryptographically chained log. Stream to Splunk, Datadog, ELK or CloudWatch. Retention up to 7 years.

Approval workflows

Multi-approver chains per environment, mandatory diff review, AI simulation before execution, auto-reject timers.

Budgets & cost

Monthly caps, per-user and per-project quotas, alerts at 50/75/90/100%, cost tags for chargeback.

Integrations

Slack & Teams, Jira & Linear, GitHub Enterprise & GitLab, PagerDuty & ServiceNow, signed webhooks.

Support & SLA

99.9% uptime, 1h response on critical issues, a named CSM and a shared Slack channel with engineering.

Compliance

Ready for your security review.

FrameworkStatusDetails
SOC 2 Type IICertifiedAnnual third-party audit. Report under NDA.
GDPRCompliantDPA available, EU data residency, erasure and portability.
CCPA / CPRACompliantDeletion and opt-out mechanisms in place.
HIPAAReadyBAA available. PHI encryption and access controls.
OWASP Top 10CompliantTested against the Top 10, with regular third-party pentests.

Security practices

  • Annual third-party penetration testing
  • CVE scanning on every build
  • SAST and secret detection in CI
  • Incident response plan, 1h P1 acknowledgment
  • Responsible disclosure program

Data protection

  • AES-256 at rest, TLS 1.3 in transit
  • Keys in AWS KMS / GCP Cloud KMS
  • Data residency: US, EU, LATAM, APAC
  • Daily backups, 30-day retention
  • Disaster recovery: RPO < 1h, RTO < 4h
Onboarding

Live in four weeks.

Week 1

Setup

SSO, security review and DPA, first projects and teams.

Week 2

Configure

Roles, integrations, coding standards, approval gates.

Week 3

Pilot

5–10 developers on real work, with live training.

Week 4

Roll out

The whole team, budgets set, first business review booked.

FAQ

Questions we always get.

Can we run Orquesta inside our own VPC?

Yes. Dedicated VPC on AWS, GCP and Azure, plus fully on-prem and air-gapped installs. Your data never leaves your cloud account.

Can we use our own models?

Yes. On-prem and air-gapped deployments can point at your own endpoints — Ollama, vLLM or any OpenAI-compatible API — so nothing leaves your network.

Can our clients prompt the agent too?

Yes, with a role. Scope what they can see and ask per project and environment, and require approval before anything they request reaches production.

What is the minimum team size?

Enterprise is built for 10+ developers, but we work with smaller teams that have specific compliance or deployment needs.

How long does SSO take?

Usually 1–2 hours for Okta or Entra ID, handled by our team during onboarding. SCIM adds another 1–2 hours.

Can we get the SOC 2 report, or a BAA?

Yes to both. The SOC 2 Type II report is shared under NDA, and we sign BAAs for healthcare customers.

What happens after the 30-day pilot?

Continue on your enterprise plan with no migration, or we delete your data within 30 days. No lock-in.

Start a pilot in one call.

30 days with a dedicated engineer. If it doesn't ship real PRs by week two, we refund the call.

oscar@getorquesta.com · SOC 2 Type II on file · DPA ready
Feedback