A complete path
Can someone finish the task, including the less convenient steps?
Why Little Hill
The visible parts of an application are only part of the job. We care just as much about how it behaves, how it holds together, and how it can evolve.
Experience
More than 20 years of professional development experience, with data at the heart of virtually every project.
That includes relational database design, SQL and Oracle PL/SQL, enterprise applications, reporting, analytics, and the business rules that connect them.
This background changes how we approach a new application. We think about what the information means, how it relates, where it comes from, and what needs to remain true as the system changes.
Discipline
AI can accelerate implementation. It doesn’t remove the need to understand what is being built.
Requirements can drift. A local fix can introduce a problem elsewhere. Without clear boundaries, business logic becomes scattered and architecture becomes harder to reason about.
We use AI tools within a deliberate process: explicit requirements, cohesive components, reviewable changes, and verification of the behaviour that matters.
Read about AI-assisted developmentFunctional quality
A successful demonstration is a milestone. It isn’t the definition of finished.
We have a low tolerance for workflows that stop halfway, interfaces that leave people guessing, and irritating defects that persist because they seem small.
No responsible developer can promise literally bug-free software. We can make correctness, completeness, and maintainability central to the work, rather than treating them as finishing touches.
What we look for
A practical standard for useful software.
Can someone finish the task, including the less convenient steps?
Do calculations, permissions, and business rules match the requirements?
What happens with missing information, invalid inputs, or an interrupted action?
Are the language, controls, and feedback consistent and understandable?
Can the application be understood and extended without working around itself?
Have the important behaviours been checked, rather than assumed?
Start with a conversation
Tell us about the problem, the people using it, and what a good result would look like.