Technical
The DevOps Minimum Viable Stack for a Solo Consultant
Solo consultants either over-invest in DevOps or ignore it. Year two I found the minimum viable stack that keeps client projects safe without becoming a second job. Here is what is in it and what is not.
The Stack
- Deployments: platform-as-a-service (Vercel, Railway, Azure App Service)
- Secrets: platform environment variables plus 1Password for personal
- Monitoring: Sentry on frontend, platform logs on backend
- Backups: automated, weekly, offsite, tested monthly
- Uptime: a single external monitor hitting
/health - Domain: Route 53 for everything I touch now
Six items. Each takes under an hour to set up. Together they cover 90 percent of the real risks.
What I Cut
- Kubernetes (almost never worth it at solo scale)
- Custom CI pipelines (GitHub Actions defaults suffice)
- Dedicated monitoring stack (Sentry and platform logs are enough)
- Multiple staging environments (preview deploys replace them)
Every one of these is worth it at larger scale. None are worth it for solo or small team consulting.
The Health Endpoint
Every app I deploy has a /health endpoint that checks critical dependencies. The external uptime monitor pings it every five minutes. One number of truth about whether the app is up.
@app.get('/health')
def health():
return {
'status': 'ok',
'db': check_db(),
'version': VERSION,
}Simple, honest, catches 90 percent of outages before clients notice.
The Backup Reality
Backups that are not tested are not backups. I keep a monthly reminder to restore one database and run a sanity check. Has caught silent backup failures twice in the past year. Each time would have been catastrophic if discovered during a real incident.
The Cost Ceiling
The whole stack costs about forty dollars a month per client project. Lower if the project is small, higher only if the client has unusual needs. Clients stop asking about DevOps cost when you can tell them the exact line items.
When to Grow the Stack
Add a piece when a specific incident justifies it. Not before. Most DevOps over-investment is pre-paying for scenarios that never arrive. Reactive additions map to real risks. Proactive additions map to other consultants blog posts.
Gregor Hohpe's architect elevator writing frames the solo-scale tradeoffs well.
The Incident Playbook
I also keep a one-page incident playbook per project. Where to look when an alert fires, who to call, where the backups live, how to roll back. Two minutes to write after a project ships, saves an hour of panic during the eventual incident. The playbook is the other half of the DevOps stack that nobody talks about.
The Retainer Benefit
Clients on retainer get proactive DevOps: monthly updates, patch reviews, backup tests. Clients on fixed-fee projects get the stack set up at launch and documentation for handoff. The difference is deliberate. Proactive operations is a retainer product, not a launch deliverable.
What I Tell Other Consultants
Resist the DevOps industrial complex. Your clients are small. Their infrastructure should be small too. The best DevOps setup is the one you can explain to a client in five minutes and hand off to their IT person in an hour.
RELATED READING
The Consulting Shift I Am Making In Year Two
After a year of writing and building, my consulting practice is changing shape. Shorter engagements. Sharper outcomes.
ReadThe Frontend Shift: Shipping Less JavaScript In Year Two
A year ago I reached for Next.js for everything. This year I often reach for nothing.
ReadThe Serverless Lesson I Would Write On A Sticky Note
After a year of shipping serverless projects, one rule explains most of the wins and all of the losses.
Read