Roadmap

What we're building next. Each track lists the goal and the first concrete slice we plan to ship.

• Managed databases
  Attach a Postgres or Redis instance to a service in one click — connection strings injected as env vars, backups and TLS handled.
  First: Postgres add-on with automatic DATABASE_URL injection and daily snapshots.
• Multi-cloud high availability
  Survive the loss of a single cloud provider. Run the control plane and tenant workloads across more than one region and more than one cloud.
  First: Stand up a second region on a different provider and replicate platform Postgres for control-plane failover.
• Billing
  Self-serve plans, automated invoicing, and credit-card-backed accounts in USD.
  First: Stripe-backed monthly invoices with usage breakdown (CPU, RAM, egress) and plan-cap enforcement.
• Referrals
  Reward users who bring other users — credits for the referrer and a head start for the referred account.
  First: Signup with a referral code grants credits to both sides on first paid usage.
• Better logging
  Searchable, retained build and runtime logs across the dashboard, CLI, and MCP — not just the last few minutes of stdout.
  First: Loki-backed retention with per-service search and time-range filters in the dashboard.
• Better metrics
  Per-service CPU, RAM, request, and egress charts in the dashboard with sane defaults and alert hooks.
  First: Per-service metrics tab driven by the existing Prometheus stack — CPU, memory, and request rate over the last 24h.
• Internal networking
  Service-to-service traffic inside a project over private DNS — no public hop, no extra auth, lower latency.
  First: Stable internal hostnames per service (e.g. `api.internal`) reachable from sibling services in the same project.
• Regions
  Let users pick where their app runs for latency, compliance, and data residency.
  First: A second region selectable at project creation, with the existing region kept as the default.
• Volumes
  Attach persistent storage to a service so stateful workloads survive restarts and redeploys.
  First: Per-service block volume mounted at a configurable path, with size set at attach time.
• Replicas
  Run more than one instance of a service for throughput and resilience, with traffic load-balanced across them.
  First: Set a fixed replica count per service in the dashboard with rolling restarts on redeploy.