Hire a fractional DevOps and IT team, not another full-time seat.
Digital Buttes works as your part-time DevOps and IT department. We build and run cloud infrastructure for software teams of 5-20 developers, on AWS and Google Cloud, for an agreed number of hours a week. You get a team's worth of experience without a senior salary, a months-long search, or one person who is the only one who knows how production works.
When fractional makes more sense than hiring
A full-time platform engineer is the right call once there is a full-time platform job. Before that, the work comes in bursts: a migration, an audit, an outage, a quarter of quiet. Fractional fits the bursts.
- You have one DevOps engineer, or none, and infrastructure work lands on whichever developer is nearest
- Hiring a senior platform engineer would take months and cost more than the work justifies yet
- You need Kubernetes, CI/CD, or real observability done once, properly, by people who have done it before
- Someone has to carry the pager and nobody wants to be the only one
- IT work such as accounts, laptops, SSO, and offboarding is eating engineering time
- A customer or auditor is asking for SOC 2 or HIPAA and the infrastructure controls are on the list
One team, three kinds of work
Most teams start with one of these and add the others as they need them. The same people do all three, so nothing falls between two vendors.
DevOps and platform
- Kubernetes on EKS and GKE
- CI/CD pipelines and GitOps
- Terraform and infrastructure as code
- Monitoring, logging, and tracing
- Cloud cost review
- Migrations and upgrades
Operations and on-call
- On-call and incident response
- Runbooks and blameless post-mortems
- Backups and restore testing
- Patching and cluster upgrades
- Database administration
- Security and compliance controls
IT operations
- Onboarding and offboarding
- Account and access provisioning
- SSO and 2FA enforcement
- Device management
- SaaS administration
- Documented SOPs
How a fractional engagement runs
-
1
Intro call and scoping
We look at what you have, what is hurting, and who is currently doing the work. We agree on hours per week and the first priorities before anything is signed.
-
2
First 30 days
Access, an inventory of what is actually running, and a quick win: the deploy nobody likes doing, the alert nobody trusts, or the thing that only one person knows how to fix.
-
3
Steady state
A shared channel, a queue of work, and a weekly rhythm. If you want us on-call, we join the rotation. Everything we build is documented and lives in your accounts and your repos.
-
4
Scale up or down
More hours during a migration or an audit, fewer once your developers own the day-to-day. There is no lock-in: if you hire a full-time engineer, we hand off to them and help them get started.
Fractional team, full-time hire, or an MSP?
Full-time hire
Right when there is a full-time platform job and budget for a senior salary. Slow to fill, and a single person means a single point of failure: when they take vacation or leave, the knowledge goes with them.
Managed service provider
Good at laptops, email, and help desk tickets. Rarely comfortable in a Kubernetes cluster or a Terraform repo, and rarely willing to be paged for your production app.
Fractional team
A small group that does both the DevOps work and the IT work, for the hours you need. Coverage when someone is out, experience from other teams like yours, and a hand-off built in from the start.
Common questions
How many hours a week?
Build engagements are typically 10-20 hours a week. Operations retainers vary with what you want covered. We agree on a number up front and adjust it as the work changes.
Do you take on-call?
Yes, for teams on a managed operations retainer. We join the rotation, respond to incidents, and write the post-mortem afterward.
Will you work with our existing engineers?
That is the point. We pair with your developers, document as we go, and leave them able to change the platform without calling us.
Which clouds and tools?
AWS and Google Cloud primarily. Kubernetes wherever it runs, plus Linux and bare-metal servers. Terraform, ArgoCD, Prometheus, Grafana, Datadog, and whatever you already have that works.
Do you handle IT as well as DevOps?
Yes. Accounts, SSO and 2FA, device management, onboarding and offboarding, and the documented procedures that keep those from becoming a weekly interruption for engineering.
Are we locked in?
No. Everything is in your accounts, in code, and documented. Scale down or stop whenever your team owns it. Our goal is that the hand-off is uneventful.
Talk to us about a fractional engagement
Tell us what is landing on your developers today. We'll tell you what we would take off their plate first.