Finding an appropriate business model is situational and may evolve as you engage with the user community. You could start by working in public with an open source license [1]. Your initial focus is customer discovery and building a community of stakeholders willing to invest their time and accepting the adoption risk. Once you have advocates, they may have suggestions for an appropriate business model. For example, the existing non-profit may decide to apply for grants, keeping the work open source and acting as the fiscal sponsor for the project via open collective [2] supporting a paid position. In my experience, getting caught up on recapturing early investment, being concerned about secrecy / proprietary value, or being worried about a competitor are often not the best use of mental energy early on. Conversely, sharing more about your space and ideas may win you collaborators and potential customers you didn't anticipate. Alas, none of this advice may apply, the right approach is specific to a particular opportunity. Good luck & have fun.
Sid, Thank you for funding Arden Bio. Rheumatology patients and their clinicians are often left to throwing darts from afar, in series and often with significant & permanent function loss with each dart. Moreover, proper care of refractory rare diseases is not something that insurance funds: my otherwise excellent coverage expressly denies "investigatory" treatments, for example. As a small fiber neuropathic vasculitus patient with secondary Sjögren's I really look forward to progress in this field, it cannot happen soon enough. Best of luck in your own care and profile. - Clark
Heath Brothers' treatise, Switch: How to Change Things When Change Is Hard, describes three legs to facilitating a change: clear vision, sufficient motivation, and concrete first steps.
The title of this blog post hints at an important topic. However, I think even a single page summary or graphic from Switch may be more actionable. I don't love Switch's elephant analogy but it's good enough. It helped me with blind spots in my proposals.
Democracy is successful when it creates the business-regulatory environment and marketplace that let the private sector advance human welfare as well as technology.
In many one-off applications there's no real need to create a separate API, tight coupling lets you focus on an MVP without having to focus on abstracting out an intermediate model -- you could go from the SQL database right to HTML rendering.
[1] https://press.stripe.com/working-in-public [2] https://opencollective.com/become-a-host