LIVESave the World, live from Salt Lake City during KubeCon · Wed Nov 11, 11:00 MT →
// the platform
Everything in the factory.
Infrared 0.1, area by area, in the order the app lays it out.
Each area links to its page in the docs, which describe every screen, command and setting. If you'd rather see it, book a call and we'll show you on real Products.
01 / productsProducts and environmentsThe repos you build, release and change together, and where they run.
- Start from a zip or a template, or set up delivery for repos you have
- Environments in promotion order, each with live pods, logs, history and settings
- Guarded or fast deploys, editable smoke and guard checks
- Custom domains with DNS checks, redirects, environment variables and secrets
- Umbrella Products: several repos shipped as one versioned chart
02 / observabilityHealth, built into every ProductNot a separate tool: health sits on the Product, where you look anyway.
- Uptime, edge error and edge latency SLOs by default
- Alerts that open incidents
- A public or signed-in status page with 30 days of uptime
- DORA figures across every Product
03 / changesChanges and the change flowEvery request from the ask to production, and everything that happened on the way.
- Ask in a sentence; the architect plans it; you triage it
- Verdict cards, evidence and logs for every step
- Agents that ask you questions, and runs you can rescue
- Every entry: the full log of runs, merges, approvals, deployments and rollbacks
04 / deploymentsDeploymentsOne version on its way through its environments.
- Every push to the default branch deploys
- Zone by zone, under a release guard with automatic rollback
- Production behind a person's approval; Go now to end a hold
- Timings: where the lead time went
05 / releasesReleasesThe versions that reached production, gathered for people to read.
- A nudge when there's enough to announce
- Release tracks, notes for engineers and a post for everyone else
- Tags, the production image and the GitHub release, done for you
06 / workforceWorkforce and agentsThe people and agents doing the work, on Teams.
- Teams, people, agents and access groups in one place
- A default catalog of AgentRoles and AgentWorkflows, all editable
- Model tiers, budgets and a circuit breaker per AgentWorkflow
- Schedules and triggers: on an issue, a label, CI or a cron
07 / clustersClusters and cloudsThe infrastructure your Products run on, built and kept in gitops.
- GKE and EKS workload clusters from templates, with OpenTofu
- Cloud accounts for Google Cloud and AWS with no stored keys
- Any other cluster registered by its kubeconfig
- Clean deletion that shows what it removes and costs first
08 / usageUsage and costWhat the agents spend, where, and against which budget.
- Spend per step, per Change, per day, per model and AgentRole
- Budgets with what's left, and rate limits before they bite
- A chip on every page with today's spend
09 / accessAccess and identityWho gets in, with which role, and how you get in when everything else fails.
- Sign in with Google or GitHub; sign-up for orgs of their own
- Access groups from Google Workspace groups, email or GitHub logins
- GitHub teams kept in step; linked accounts with no stored forge tokens
- Break glass that's recorded, and API tokens with their owner's live roles
10 / settings and operationsSettings and running an installEverything an org or an install needs, on one page, and the jobs of keeping it running.
- Model provider, Slack, summary videos, domains
- A read-only MCP server and a code index for agents
- Encrypted backups, restore and guarded upgrades
- The ir CLI and an API reference served by every install
See it on your Products.
A tour of the platform is better on your own code. Book a call and we'll take one of your issues through it.