Pay for the work done.
PETROS and Tiphys are AI teammates: they take the repetitive work, and your team stays in charge. You pay for what they do, by cloud account and by cluster. Start with a 30-day pilot, credited in full when you subscribe.
PETROS
Platform engineering legwork for your cloud, with your team approving every change. Priced by cloud account.
Pilot
Thirty days on one of your AWS accounts, to see the work it does.
- 1 cloud account
- 150 tasks
- Credited in full toward your first payment if you subscribe within 30 days
Team
Most teamsFor platform teams running production on AWS.
- 3 cloud accounts included, then $249 per additional account
- 500 tasks a month
- GitHub, GitLab and Argo CD
- Terraform Cloud
- Kubernetes: control plane, nodes and add-ons
- Datadog, Prometheus + Grafana and PagerDuty
- Jira and Slack
- Approval gates
- Priority support, within 1 business day
Enterprise
One agreement for every product, for organizations with many teams.
- Annual agreement
- Custom task and investigation pools
- Your own models: prompts and data stay in your AWS account
- SSO/SAML and the audit log
- ServiceNow
- Custom runbooks and policies
- A named contact and an SLA
Hours handed back to your team.
Put in your own numbers. PETROS does the legwork on each task, your engineers review and approve, and the hours saved go to the higher-value work.
These are the defaults. Turn on JavaScript to try your own numbers.
- Hours handed back to your team each month
- 625500 tasks × 1.25 h each, after review
- What those hours cost today
- $62,500625 h × $100 an hour
- Team plan for that volume
- $999 a monthTeam, up to 3 cloud accounts, before any bundle or annual discount
An estimate from the numbers you enter, not a quote. People review and approve every change PETROS proposes, and that review time comes off the hours above. The hours handed back go to the work your team never quite gets to.
Tiphys
SRE legwork for your workloads, with your team approving every write. Priced per cluster.
Pilot
Thirty days on one cluster, on your real workloads.
- 1 cluster
- 50 investigations
- Credited in full toward your first payment if you subscribe within 30 days
Team
Most teamsFor teams running workloads on several clusters.
- 40 investigations per cluster a month, pooled
- Remediation pull requests
- Datadog, Prometheus + Grafana and PagerDuty
- Slack and Jira
- Approval workflow
- Priority support
Enterprise
One agreement for every product, for organizations with many teams.
- Annual agreement
- Custom task and investigation pools
- Your own models: prompts and data stay in your AWS account
- SSO/SAML and the audit log
- ServiceNow
- Custom runbooks and policies
- A named contact and an SLA
Hours handed back to your team.
Put in your own numbers. Tiphys does the legwork on each investigation, your operators review and approve, and the hours saved go to the higher-value work.
These are the defaults. Turn on JavaScript to try your own numbers.
- Hours handed back to your team each month
- 7040 investigations × 1.75 h each, after review
- What those hours cost today
- $7,00070 h × $100 an hour
- Team plan for that volume
- $399 a monthTeam on 1 cluster, before any bundle or annual discount
An estimate from the numbers you enter, not a quote. People review and approve every change Tiphys proposes, and that review time comes off the hours above. The hours handed back go to the work your team never quite gets to.
What you pay for
- Cloud account
- One AWS account. GCP and Azure later.
- Task (PETROS)
- One unit of agent work: a finding investigated, a pull request opened, or a ticket filed or updated. Scheduled scans don’t count.
- Investigation (Tiphys)
- One workload issue diagnosed end to end, whether an alert or a person in Slack started it.
Prices in USD. A pilot is paid once; Team is billed monthly or yearly; Enterprise is an annual agreement.