How we work
From first call to working software
Most client relationships follow a progression: diagnose, then fix, then build. Each step can stand alone. You are never locked into the next one.
We start with the data model, not the software. Most engagements arrive in one of three shapes: a prototype that needs a production foundation, a spreadsheet the business has outgrown, or off-the-shelf tools that no longer talk to each other.
The first hospital systems we built in 2012 are still in production. 30+ engagements have followed across hospitals, manufacturing, field operations, education, and finance. See how a Victorian health service automated its regulatory reporting and how a Melbourne wealth practice unified adviser, client and platform data in one CRM.
Once you know which problem to solve, our three engagement tiers set the scope and price up front.
Four steps. Each one optional.
Step 1: Discovery call
Free, 30 minutesA 30-minute conversation. You describe the problem. We tell you honestly whether we can help. No pitch. No proposal. Just a straight answer.
Step 2: Systems diagnostic
2-3 weeks, fixed priceWe map your current systems, data flows, and manual processes. We find the waste. You get a prioritised list of what to fix, with a clear plan for the top three items.
Step 3: Recommendations and decision
You decide what to act onThe diagnostic gives you three categories: things you can fix yourself, integrations we can build for you, and custom software if needed. You decide what to act on, and we agree scope and cost up front for whatever comes next.
Step 4: Build and deliver
Working software every 2-4 weeksWorking software every 2-4 weeks. Not a 6-month disappearing act. You see progress continuously and can change direction based on what you learn.
What if the diagnostic doesn't find anything?
If we can't find meaningful improvements, you walk away with a clean bill of health for your systems.
Since 2007, this hasn't happened.
The diagnostic does not commit you to further work with Devinium. Some firms take the report and act on it themselves. That's fine.
They have brought to the table key knowledge and competences allowing us to transition and integrate everything into our corporate environment with minimal client disruption and zero downtime.
Common questions
We tried something like this before and it failed.
Most technology projects fail because they start with the software instead of the data. We start with the data model. The hospital systems we built in 2012 are still in production.
My team won't use new software.
We design interfaces for the people who actually use the system. If your team doesn't adopt it, we've failed. That's why we deliver working software every 2-4 weeks. Your team tests it early, not after 6 months of building the wrong thing.
We're too busy for a big project right now.
The diagnostic takes about 3 hours of your team's time over 2-3 weeks. We do the rest. You don't need to clear your calendar.
How is this different from hiring a freelancer?
Continuous operation since 2007. A small senior team. We'll be here for phase 2, the support call, the 'add this feature' request in 3 years. A freelancer might not be. More about the team and how Devinium got here.
Can't we just use Zapier / Power Automate / off-the-shelf?
Sometimes yes. The diagnostic will tell you when off-the-shelf fits and when it doesn't. We don't build custom software when a cheap subscription tool will do the job. But when you hit the limits of no-code, when performance, security, or complex logic matter, that's when a proper data foundation pays for itself.
We've been prototyping with Claude, Cursor, or Lovable. Can you take it to production?
Yes. The prototype itself is a real asset. It proves the idea and the screens your team actually wants, and we keep that part. What usually has to change underneath: a proper data model, because most AI-generated code improvises a shape that will not survive a second user. Schema migrations you can run safely. Real authentication rather than a hardcoded login. An audit trail. Error handling that surfaces failures instead of hiding them, and monitoring that tells you when things break silently. We port the parts that already work and rebuild the foundation around a real database. Screens mostly stay. The system underneath becomes something you can put in front of your team or your customers without holding your breath.
Start with a conversation.
A 30-minute discovery call costs nothing. We'll tell you honestly whether we can help.