Software that keeps working at 2am on a Tuesday.
Monthly care plans covering bug fixes, security patches, monitoring and minor enhancements, with an AI add-on for evals, cost control and model updates.
What is application maintenance and support?
Application maintenance and support keeps software secure, current and working after launch: bug fixes, security patching, dependency updates, monitoring and minor enhancements under an SLA. Eazyware's care plans add an AI layer, evaluation regression when models change, prompt tuning, cost optimisation and re-indexing, in three monthly tiers.
| Service line | Support |
|---|---|
| Engagement | Monthly retainer |
| Duration | Monthly, 30-day notice |
| Starting price | $1,000 / mo |
| Typical range | From $1,000 / month |
| Deliverables | 4 listed below |
| Delivered from | Bengaluru, India (IST, UK and US East hours) |
| Code ownership | Client owns code, infrastructure, prompts and documentation |
What problem does it solve?
Delivery ends; entropy doesn't. Dependencies age, APIs change, models get deprecated, costs creep. Without ownership, the system you paid for degrades.
How do we approach it?
Care Plans begin with an onboarding audit so we know what we are looking after: architecture, dependencies, monitoring, backups, access and, for AI systems, prompts, models, evals and cost. Monitoring and alerting are set up or verified, a named support channel is opened, and a monthly cadence starts: patching, dependency updates, security checks, a health report and the enhancement hours you have chosen. For AI systems the add-on runs eval regression on every model or provider change, tunes prompts and routing for cost, re-indexes retrieval and watches for drift. Quarterly reviews look at what should change next.
What do clients use it for?
- Ongoing care for systems we built
- Taking over an application after an audit
- AI model and prompt care as providers change
- Cost optimisation for inference-heavy products
Is it the right fit?
Good fit when
- Any production system without an in-house owner
- AI products that need evals kept current
- Companies wanting predictable monthly support
Probably not when
- Projects still in active build
- Systems with no monitoring access
What do we build?
- Corrective: bug fixes, incident response, L2/L3 support
- Preventive: dependency and security patching, backups, database maintenance
- Adaptive: third-party API changes, OS and browser updates, compliance updates
- Enhancements: minor features within the monthly hour bank
- Monitoring, uptime alerts, monthly health report
- AI add-on: eval regression on model updates, prompt tuning, cost optimisation, re-indexing, drift checks
What you get
- Named support channel
- Monthly health report
- Patch and release log
- SLA tracking
How does the engagement work?
- 01
Onboarding audit
- 02
Monitoring setup
- 03
Monthly cadence
- 04
Quarterly review
What does good look like?
Software that keeps working at 2am on a Tuesday, with incidents handled inside the SLA and a report each month that tells you what was patched, what broke and what it cost. AI systems whose accuracy and cost are tracked over time rather than discovered to have drifted. Enhancement hours that get small things done without a project. And a team that knows your system well enough to advise on the next build.
How does it compare?
| Eazyware | Typical agency | In-house hire | |
|---|---|---|---|
| Time to first result | Sprint Zero in 10 days, then a fixed-scope build | 6–12 weeks of discovery before a proposal | 3–6 months to hire, then ramp |
| Pricing model | Monthly, 30-day notice | Time and materials, open-ended | Salaries, tooling, management overhead |
| AI depth | Multi-model, evals, cost routing, observability as standard | Often a single vendor API and a prompt | Depends entirely on who you can hire |
| Ownership | Client owns code, infra, prompts and docs | Sometimes retained or licensed back | Owned, but concentrated in one or two people |
| After launch | Care Plans with SLA and AI add-on | Change requests at hourly rates | Ongoing headcount whether or not there is work |
Which pitfalls do we design around?
Maintenance goes wrong when nobody owns it, when patches are skipped until something breaks, when AI systems are treated like static software and degrade as models change, and when support is a ticket queue with no context. A named engineer, a monthly cadence, AI-specific care and an onboarding audit are the answers.
What do we measure?
Every engagement is instrumented. These are the numbers you see in the dashboard and the monthly report, not claims on a website.
- SLA response and resolution times met
- Uptime
- Inference cost trend
- Eval regression caught before release
Which technologies do we use?
- Your stack
- Uptime and error monitoring
- Langfuse for AI systems
Who does the work?
A named support engineer, an on-call rotation matched to your tier, and for the AI add-on, an AI engineer who runs the evaluation and cost reviews.
What do you need to bring?
Access to the codebase, infrastructure, monitoring and, for AI systems, model provider accounts and eval sets. A named contact on your side, the escalation path for incidents, and a view on what the enhancement hours should go to first.
Frequently asked questions
Systems you didn't build?
Yes, after a paid audit of two to three weeks.
Unused hours?
Roll over one month.
Contract?
Monthly with 30-day notice. Annual gets two months free.
Where does this fit?
Software Maintenance & Support is part of our Support line. Not sure yet? Start with Sprint Zero, a ten-day discovery whose fee is credited to this build. See all pricing or talk to an engineer.