Software Maintenance & Support in Rwanda

What is software maintenance and support, and who needs it

It’s an ongoing service with agreed priorities, response times and scope, covering corrective fixes, preventive updates and small adaptive changes to your software and the environment it runs in. It suits organisations whose systems were built by a developer who is no longer available, those running software that hasn’t been updated since launch, and any organisation whose operations depend on a custom system.

The problem it solves

Software decays even when nobody touches it. Libraries develop known vulnerabilities, browsers and phones change, certificates expire, hosting platforms retire old versions, and small bugs accumulate into workarounds. Knowledge of how the system works leaves with the developer. Regular maintenance keeps the system safe and current, and documentation keeps it maintainable by someone other than its author.

What maintenance covers

Corrective. Bugs fixed by priority, with root causes recorded.

Security. Dependencies and frameworks updated, vulnerabilities patched and secrets rotated.

Environment. Hosting, database and operating system updates, and certificate renewals.

Compatibility. Changes in browsers, mobile operating systems and app store rules handled.

Backups and recovery. Backups checked and restores tested.

Monitoring. Errors, performance and availability watched, with alerts.

Small improvements, within an agreed monthly allowance.

Documentation kept current with every change.

Taking over from another developer

Many maintenance agreements start with a handover from someone else. The first weeks focus on access and understanding: securing the code repository, hosting and database credentials in your name, building a working test environment, mapping dependencies and their security status, and documenting how the system is deployed. Urgent risks found during that review, such as unsupported frameworks, missing backups or exposed secrets, are fixed first.

Priorities that match the business

• Critical: the system is down, data is exposed or a key process has stopped. Fastest response, worked until resolved.
• High: a main function fails for a group of users, with no workaround.
• Normal: a problem with a workaround, or affecting one user.
• Request: questions, small changes and improvements, planned into the monthly allowance.

Response targets for each level are written into the agreement, and monthly reports show how they were met.

Releases without surprises

Every change, however small, goes through the same path: made in version control, tested in a separate environment, approved, released at an agreed time, and recorded with a way to roll back. That discipline is what lets a system be changed for years without becoming fragile.

What the work involves

1. Review the system, code, hosting, dependencies and documentation.
2. Fix urgent risks found in the review.
3. Agree priorities, response times and monthly allowance.
4. Set up monitoring, a test environment and a release process.
5. Maintain on a regular cycle, with monthly reports.
6. Plan upgrades and improvements each quarter.

What’s included and what’s not

Typical timelines and what drives cost

Onboarding a system we didn’t build typically takes one to three weeks, depending on its condition. Maintenance is a recurring fee driven by the system’s size and complexity, the number of integrations, the age of its technology, response targets, coverage hours and the monthly improvement allowance.

Equipment typically used

Version control, issue tracking, automated tests, dependency scanning for known vulnerabilities, separate test and live environments, error and performance monitoring, and backup and restore tooling.

Common mistakes we see

No maintenance after launch. Code and hosting held only by the developer. Dependencies never updated. Fixes made directly on the live system. No test environment. Backups never restored. And upgrades postponed until a component reaches end of support.