Software Maintenance and Support Services

The four types are corrective, adaptive, perfective, and preventive. Corrective fixes defects. Adaptive adjusts software to changes in operating systems, APIs, and infrastructure. Perfective improves performance and adds features. Preventive reduces future failures through refactoring and technical debt reduction. Most engagements need all four.

Maintenance is proactive and planned, scheduled improvements, upgrades, and refactoring that address non-urgent work. Support is reactive, resolving urgent incidents that disrupt operation or cause downtime. Maintenance protects the system's future; support protects today's users. A complete agreement covers both.

We staff a conventional multi-tier support (L1, L2, L3, L4) structure and agree the escalation paths before we take the wheel.

1. End-user support (L1) handles first-line questions, query tracking, configuration and account administration, and password and access issues, escalating what it cannot close — delivered by support specialists.
2. Application support (L2) covers continuous software monitoring and health checks, performance issues, log investigation, database administration, and basic infrastructure requests, delivered by application and infrastructure support engineers.
3. Development support (L3/L4) handles code-level defect resolution, new features and integrations, and complex infrastructure work, delivered by software engineers, QA engineers, DevOps engineers, cloud platform specialists, and external vendor engineers where a third-party platform is involved.

Every incident is classified by business impact.

• P1 - critical outage means the system is down or unusable.
• P2 - major function broken means a key feature is failing and business operations are affected.
• P3 - minor defect means a workaround exists and the issue is not business-critical.
• P4 - cosmetic or request covers UI issues and improvement suggestions.

System monitoring and incident management runs underneath every support tier: proactive alerting on threshold breaches, 24 × 7 monitoring coverage where your system warrants it, and a documented incident lead for every P1.

SLAs are agreed per engagement rather than sold as a fixed package, and reported against monthly, so SLA compliance is measurable rather than asserted. Response and resolution targets are set per severity level during onboarding, against your system's availability requirements and coverage hours. Coverage scales with your needs, scalable ongoing support means adding or removing tiers without renegotiating the whole contract.

Saigon Technology's listed rate for implementation work is $22–$46/hour with senior oversight. Annual support and maintenance costs typically run 15–20% of original build cost. Actual pricing depends on software complexity, coverage hours, change frequency, number of users, and documentation quality, the main drivers of total cost of ownership.

Yes - most of our maintenance engagements start that way. We run a structured knowledge transfer producing a dependency map, a documented defect and risk register, and deployment runbooks before taking responsibility. On the ARB Australia engagement we inherited an unmaintained codebase and stabilized it; on Loan City we retained the original engineers through the transition.

It depends on your sourcing model constraints. In-house software maintenance gives maximum control but requires hiring for legacy and niche skills and carries fixed cost through quiet periods. Outsourced software maintenance ramps faster and scales both ways. Many clients choose in-house management with outsourced resources; others move to fully outsourced maintenance and support including infrastructure maintenance and support.

Reporting covers SLA attainment, incident volume and trends, root cause patterns, preventive fixes delivered, and change throughput. Do not publish specific figures or frequencies until delivery confirms what we commit to contractually.