Product map

Everything the app needs, in one operating contract.

Deploy the application first. Add only the backend, observation, and operating capabilities it needs. The dashboard, API, CLI, and MCP server share the public operation contract. The supported declarative subset uses the same resource model.

Ship an app

Build

Turn a repository into a deployed application and give every change a real place to run.

App runtime

Build and deploy a repository to a stable URL with environment variables, domains, deployment history, and rollback.

Preview environments

Create an isolated application environment for a pull request and remove it when the review ends.

Backend

Add only the state, identity, delivery, and application services that the product needs.

Database and recovery

Provision managed Postgres, inspect it, set recovery policy, create isolated branches, and use guarded restore promotion.

App Auth and access

Configure application sign-in and sessions, manage application users, and set public or organization access policy.

Object storage

Create project or organization buckets and use short-lived, scoped upload and download grants.

Realtime

Publish channel events and issue bounded capabilities, with retained replay inside the configured window.

Queues and cron jobs

Run at-least-once background work with retries and dead letters, then schedule HTTP or queue targets by time zone.

Managed cache

Provision an environment-scoped cache and deliver its connection information to the server runtime.

AI Gateway

Use allowlisted model aliases through one project gateway with explicit limits and usage records.

Customer payments

Connect payment credentials, manage products and prices, and create hosted checkout sessions from the server runtime.

Observe

Read application behavior from product activity, real-user performance, runtime evidence, and alert state.

Web Analytics

Review privacy-safe page views, visitors, custom events, and bounded traffic breakdowns after collection is enabled.

Speed Insights

Measure real-user web vitals by route, device, percentile, and time window under an explicit collection policy.

Metrics, logs, and alerts

Query runtime metrics and structured logs, then evaluate alert policies and follow incident transitions.

Operate

Investigate change, maintain the project, and keep cost decisions inside a reviewable control plane.

Project operator

Keep durable investigation threads, evidence, proposed actions, human decisions, and verified results together.

Maintenance

Analyze defined maintenance categories, review version-bound actions, and record the result of each run.

Usage and budgets

Read metered usage and set a project budget that alerts by default or blocks new deploys when its hard stop is enabled.

Give your agent a cloud it can operate.

Ship an app

No free tier. No seats. Pay for what you run.