Keep your software running and improving

Software maintenance and support is everything that keeps your software working well after launch. It falls into four kinds of work, and a good plan covers all four.

Fixing: bugs and problems found by your staff, your customers or monitoring. Keeping up: changes forced from outside, such as a new phone operating system, a browser update, a payment gateway changing its API or a new tax rule. Preventing: security updates, monitoring and backups that stop small problems turning into outages. Improving: the small changes your team asks for that make the software fit the business better each month.

Support is how you reach the people who do this work, and how quickly and clearly they keep you informed. It should feel like having your own technical team, without hiring one.

Software doesn’t wear out, but everything around it changes. Apple and Google release new versions of iOS and Android every year, and app stores require apps to keep up. Programming languages and frameworks stop receiving security fixes after a few years. Certificates and domains expire. Payment gateways, couriers and government portals change how they work.

The rules change too. GST rates and e-invoicing requirements shift, payroll ceilings for PF and ESI are revised, the RBI updates its directions, and privacy laws such as India’s DPDP Act and Europe’s GDPR set new duties.

And your business changes: new products, new branches, new staff, new questions from customers. Software that isn’t looked after slowly stops fitting, and then one day stops working.

Software often outlives the people who built it. When we take it over, we start with a review: the code, how it is deployed, the servers and cloud accounts, domains, app store and payment accounts, backups and whatever documentation exists.

The review usually finds the same things: backups nobody has restored, admin passwords shared by email, accounts in a former developer’s name, versions that no longer get security fixes and no record of how to release a change. You get a written list of these risks, most serious first.

We fix the urgent ones before anything new: accounts moved into your company’s name, passwords changed and two-factor switched on, a tested backup, monitoring, and a documented way to release. Then we write down how the system works, so knowledge stays with your business.

A good fix removes the cause, not just the symptom. Every issue is logged in a tracker you can see, reproduced on a test copy, traced to its cause, fixed, tested and released, and where it makes sense a test is added so it cannot return unnoticed.

Priorities are agreed with you. A problem that stops sales, payments or your team’s work goes first; a typo can wait for the next release. You can see where every issue stands at any time, without having to chase anyone for an update.

Reporting is simple. Your team writes to [email protected], adds it to your own tracker or sends a WhatsApp message, ideally with a screenshot and what they were trying to do. We turn it into a proper issue with the steps to repeat it, so nobody has to explain the problem twice.

Many security problems come from software that simply wasn’t updated. We keep frameworks, libraries, servers and databases on versions that still receive security fixes, using tools such as Renovate or Dependabot to raise each update, and automated tests to check it before release.

We apply updates in small steps every month rather than letting them pile up, because a jump of several versions at once is where upgrades become expensive and risky. Bigger upgrades, such as a new major version of PHP, Node.js or a framework, are planned ahead and done before the old version’s support ends.

App store rules are part of this too. Google Play and Apple’s App Store require apps to target recent versions and declare how they use data, and we keep your apps compliant so updates are never blocked.

You shouldn’t learn about an outage from a customer. We set up monitoring that checks your site and apps around the clock, records errors with tools such as Sentry, and watches the things that fail quietly: background jobs, payment webhooks, email delivery, disk space, certificates and domain renewals.

When something goes wrong, the alert reaches us, and work starts with the details already in hand. Dashboards in tools such as Grafana show trends, such as pages getting slower or errors creeping up, so they can be fixed before they become outages.

A backup only counts if it restores. We set up automatic backups of your databases and files, kept in more than one place, and restore them to a test copy on a schedule to prove they work. The result goes into the monthly report.

We also write down what happens if the worst occurs: who does what, in which order, and how long a restore takes for your system. Our managed cloud hosting partner in Mumbai, for example, keeps customers’ backups in more than one location and tests recovery regularly.

When a rule changes, your software has to be ready on the day it takes effect. We track changes that affect your system and plan them ahead: GST rates and e-invoicing, e-way bill rules, PF, ESI and professional tax, RBI directions for lenders and payments, and privacy laws in India, the UK and Europe.

Our portfolio shows this kind of work over many years. We updated a payroll portal for every budget and statutory change, including the PF wage ceiling rise in 2014; switched a coaching institute’s fee receipts from service tax to GST in 2017; added consent records and data deletion for a tour operator when the GDPR arrived in 2018; and brought FASTag toll records into a transport company’s fleet dashboard once FASTag became mandatory.

Maintenance is not only about standing still. Your staff and customers have ideas: an export button, a faster search, a new report, a field that should fill itself in. We keep them in a list with you, agree each month which to do, and release them in small, tested steps.

Small, regular improvements add up. Over a year the software fits the business better, without the risk and cost of a big rebuild. When a bigger change is needed, such as a new module or a move to the cloud, we plan it as a project of its own.

Sometimes the honest answer is that the old system has reached its limit. When every change becomes slow and risky, we say so, and plan a step-by-step modernisation instead.

Each month you receive a short report in plain words. It lists what was fixed, what was updated, what monitoring caught, the result of the backup check, what was improved, any risks ahead such as versions nearing the end of support, and the plan for next month.

The report is how you know the work is being done, and it is the basis for deciding together what comes next.

Once a year we step back further. Together we look at which versions reach the end of their support in the coming year, which rule changes are on the way, what the software costs to run and which parts slow your team down most. The result is a short plan for the year ahead, so the bigger upgrades are budgeted for rather than discovered in a hurry.