DevOps for Small Teams: Keeping Grant-Funded Software Alive After the Grant
A practical handover and operations checklist for university and NGO projects: hosting, CI/CD, backups, monitoring, security updates, documentation and ownership — so the system outlives its funding.
A familiar story in universities and NGOs: a funded project builds something good, the pilot goes well, the final report is submitted — and a year later the site is down, the domain has lapsed, or a phone update broke the app. Nobody is responsible any more because the project staff contract ended.
Sustainability is not a paragraph in the proposal. It is a set of engineering and ownership decisions made during the project. Here is the checklist we use.
1. Decide ownership on day one
- Named owner after the project ends: a person or unit, not “the project team”.
- Accounts in the institution’s name: domain, cloud account, app-store developer account, email services. Never a staff member’s personal account or credit card.
- Source code in an institutional repository (GitHub or GitLab organisation owned by the department or IT office).
Critical accounts registered to a research assistant whose contract has ended. Fix this in the first month, not the last.
2. Host where someone will still be in two years
| Option | Good for | Watch out for |
|---|---|---|
| Central university IT hosting | Long-term stability, SSO, security scans | Lead time for approval; follow their stack |
| Faculty / department server | Faculty-owned content, quick access | Who patches the OS? |
| Cloud (institution account) | Scalability, managed databases | Ongoing bills need a budget line |
| Static hosting (for content sites) | Near-zero maintenance and cost | Not for apps with logins or databases |
For many teaching and research websites, a statically generated site is the most sustainable choice: nothing to patch on the server, very little to break.
3. Automate the boring things
A minimal pipeline that pays for itself many times over:
# .github/workflows/ci.yml — illustrative
name: ci
on: [push, pull_request]
jobs:
test-and-build:
runs-on: ubuntu-latest
steps:
- uses: actions/checkout@v4
- uses: actions/setup-node@v4
with: { node-version: 22 }
- run: npm ci
- run: npm test
- run: npm run build
- Every change is built and tested automatically. Future maintainers can change things without fear.
- Deployments are scripted, not done by hand from someone’s laptop.
- Dependency updates arrive as automatic pull requests (Dependabot or Renovate) and are merged monthly.
4. Backups you have actually restored
- Daily database backups, kept off the main server, retained for at least 30 days.
- Test a restore before handover — a backup nobody has restored is a hope, not a backup.
- Write the restore steps in the runbook.
5. Monitoring that reaches a human
- Uptime checks on the main pages and login.
- Certificate-expiry alerts (HTTPS certificates still catch teams out).
- Alerts go to a shared mailbox or team channel, not one person’s inbox.
6. Security basics
- Single sign-on where possible; no shared admin passwords.
- Least-privilege roles for editors and admins.
- Regular OS and dependency patching, scheduled outside teaching peaks.
- Personal data minimised and handled under your institution’s data policies.
7. Documentation people will read
Three short documents beat one long one:
- README — what it is, how to run it locally, how to deploy.
- Runbook — what to do when it’s down, how to restore, who to call.
- Admin guide — for non-technical staff: adding users, editing content, exporting data.
8. A realistic support budget
Software needs care after launch: security patches, browser and OS changes, small fixes. Budget a modest annual amount for the first year or two, and decide who pays when the grant ends — the department, central IT, or a shared service.
Handover checklist
- All accounts and domains owned by the institution
- Source code in an institutional repository
- CI pipeline green; deployment scripted
- Backups running and a restore tested
- Monitoring and certificate alerts to a shared address
- README, runbook and admin guide delivered
- Admins trained; named owner confirmed
- Support arrangement for the next 12 months agreed
Need help taking over or stabilising a system a previous project left behind? That is a large part of what our DevOps and care plans cover.
Planning a project like this?
Gigantic Work builds web platforms, research portals, AI tools, Unity/VR experiences and cloud infrastructure for Hong Kong universities, grant-funded projects, NGOs and SMEs — and hands over everything at the end.