cMind for cloud & VPS providers 🖥️
You already rent compute. cMind is a ready-made, open-source product you can wrap that compute around: offer managed cMind hosting and land a high-value, sticky, compute-hungry workload — algorithmic traders, brokers, prop firms, and trading communities who want the platform running without becoming the ops team themselves.
Run the stateless tier + Postgres + a node fleet; hand customers a branded URL. Monetize the subscription, the compute, the white-label, and the AI. → Deploy to the cloud
Why offer managed cMind
- No build cost. It's open source, MIT-licensed, and already documented, tested, and containerized. You package and operate it — you don't build it.
- A differentiated product for a lucrative niche. Algo trading is compute-hungry: backtests and live nodes burn CPU, which is billable usage you already sell.
- Sticky customers. Traders who build and run strategies inside the platform don't churn casually.
- Turns a caveat into an upsell. cMind is self-hosted by design — for customers who "don't want to be the ops team," you are the answer.
Who buys managed cMind from you
- Individual quants & traders who want it hosted. → For traders
- cTrader brokers running a white-label for their clients. → For brokers
- Prop firms & copy-trading businesses who need branded, auditable infrastructure.
What "managed cMind" means to run
You operate three tiers; the customer gets a branded web URL:
| Tier | What it is | Where it runs |
|---|---|---|
| Stateless (Web + MCP) | The app + API + MCP server | Any container platform, autoscaled |
| Database | PostgreSQL | Managed Postgres (RDS / Flexible Server / your own) |
| Node fleet | Builds & runs cTrader containers | VMs or Kubernetes — needs privileged Docker |
Node agents build and run cTrader containers, so they need privileged Docker. That rules out serverless container runtimes (Azure Container Apps, AWS Fargate) for the agents — run those on Kubernetes, a VM, or EC2. The stateless tier runs anywhere.
Real, copy-paste deployment guides make this concrete: cloud overview · Azure · AWS · Kubernetes · Scaling.
How you monetize it
- Managed hosting subscription. Monthly Starter / Team / Business plans sized by node fleet and backtest concurrency.
- Usage & compute metering. Bill backtest-hours, live-node-hours, and storage — naturally metered by the container fleet you already run.
- White-label reseller tiers. Charge more for a full rebrand (logo, colors, PWA,
ShowSiteLink=false) and for enabling premium capabilities via feature toggles. → White-label - Managed AI. Bundle a default AI provider key so every customer's users get AI with no setup, and mark up the usage — or offer bring-your-own-key. → AI feature
- Prop-firm & copy-trading revenue share. Host firms running challenges and performance fees and take a platform cut. → Prop-firm · Performance fees · Provider marketplace
- Setup, onboarding & SLA. Attach professional services and premium support.
Multi-tenant patterns
- Deployment-per-tenant (recommended). One branded instance per customer — strong isolation,
per-tenant branding and database, a distinct node join token per tenant. Branding is read from
IOptionsMonitor, so each instance carries its own identity. → Multi-tenant branding · Node discovery - Shared control plane (advanced). Drive many instances from your own provisioning layer, seeding branding and features per tenant programmatically.
Metering usage for billing
An owner/admin-only GET /api/usage endpoint returns a read-only summary a provider can poll and
bill on — without any new domain or persistence, it projects existing state:
{
"users": { "total": 42 },
"nodes": { "total": 6, "online": 5 },
"instances": { "total": 1280, "backtestsRunning": 3, "runsRunning": 11 },
"cbots": { "total": 210 },
"tradingAccounts": { "total": 88 }
}
Poll it per tenant deployment to drive seat-based, fleet-based, or workload-based pricing. Pair with logging & observability for finer compute metering.
Keeping margins predictable
Scale nodes to demand, share Postgres tiers, and autoscale the stateless tier. The operational surfaces you need are already there:
Get started
- Stand up a reference deployment from the cloud guides.
- Template it per tenant (branding + join token + DB) and wire your billing to compute usage.
- List it — you now have a managed algo-trading platform to sell.
Contribute back
Providers running cMind at scale hit the sharp edges first. Upstreaming your operational fixes and IaC improvements keeps your fleet cheap to maintain — start with the Contributing guide.