Workflow discovery
Best when: You know the task is painful, but the right product is not yet clear.
Current-workflow mapping
User and decision points
Product opportunity and boundaries
Recommended smallest next step
Services
Start with the task. Find the smallest product that changes it.
You can begin before you have requirements or designs. AI helps us move quickly from the real problem to a working solution, then becomes part of that solution wherever it adds genuine value.
Read how to decide whether custom software is warranted
Prototype or MVP
Best when: You want to make the idea tangible and test it before a larger investment.
Focused feature definition
Rapid AI-assisted implementation
Interactive workflow or working MVP
Build roadmap based on evidence
Custom product build
Best when: The workflow is understood and you are ready for a usable web or mobile product.
Product and interaction design
Web or cross-platform development
Applied AI features where valuable
Testing, deployment, and launch support
Improve an existing product
Best when: You have a working tool, but its workflow, usability, or automation needs attention.
Product and workflow review
AI opportunity assessment
Targeted feature development
Ongoing iteration support
AI, twice applied
Faster delivery. Smarter products.
AI-accelerated development
I use AI throughout research, prototyping, coding, testing, and iteration. That shortens the path from a recurring problem to a usable product and makes focused custom work practical.
AI inside the solution
I design AI features that summarize, classify, recommend, generate, search, or automate within the real workflow—while keeping important decisions reviewable and in human hands.
Not sure which engagement fits?
Describe the task as it exists today. The first recommendation may be a prototype, a complete build, an improvement to an existing tool—or no custom software at all.
Tell me about the workflowBefore we begin
Common questions
Do I need a technical specification before getting in touch?
No. A clear description of the task that keeps getting in the way is enough. Requirements, wireframes, and technical terminology are outputs of the discovery process, not prerequisites for it.
How small can an engagement be?
The smallest useful engagement is a focused workflow conversation that ends with honest scope and feasibility feedback and a practical recommendation for the next step. That recommendation is sometimes a prototype, sometimes a full build, and sometimes no custom software at all.
What does "smallest useful product" mean in practice?
It means identifying the single part of the workflow where software changes the outcome, building that well, and putting it into real use before expanding. It is narrower than a conventional MVP because it is scoped to one decision or one recurring task rather than to a product vision.
How is AI used in the work?
In two distinct ways. AI accelerates research, design, implementation, testing, and iteration, which makes focused custom work economically practical. Separately, AI features are built into the product itself when they can summarize, classify, recommend, generate, search, or automate within the real workflow — with consequential decisions kept reviewable and in human hands.
Will you recommend against building software?
Yes, when that is the honest answer. Some workflows are better served by reconfiguring an existing tool, changing a process, or accepting a manual step. A recommendation to build nothing is a legitimate outcome of discovery.
Do you work with clients outside Hamilton and Ontario?
Yes. JW Soft is based in Hamilton, Ontario, and works with clients and collaborators remotely wherever the right problem is.
What technologies do you build with?
Web and cross-platform mobile applications, chosen to fit the workflow rather than a fixed stack. Technology selection happens after the workflow is understood, not before.
Who actually does the work?
Peter Basl. JW Soft is an independent studio, so strategy, design, development, launch, and ongoing operation are handled by the same person who ran the discovery conversation.