I am evaluating if this idea has merit. Comments and feedback welcome.
Long long ago, when I used MS Access (probably Access98), there was a functionality where all forms, master-detail views etc., could be autogenerated using a wizard. In those days, I found it to be very cool and used it for prototyping apps client-server/n-tier architecture apps.
Is there a market for enterprise CRUD apps that are completely schema-driven i.e., a data analyst/programmer specifies the data model (such as an Entity-Relationship Model) and the entire database, data validation and business logic, service layer and UI is autogenerated.
I also have some ideas around autogenerating entire business process workflows from the same data schema.
As a simple example, say we are modeling automotive insurance claims model with
- Subscriber
- Vehicle
- Policy
- Claim
entities. The business process flows might be:
- *Subscriber onboarding*: This adds a new subscriber , new policy and the vehicle on the policy
- *Policy servicing*: This adds vehicles or policy riders
- *Claim submission and processing*: This adds a new Claim record for a subscriber, vehicle and policy that are matched against the policy information for further settlement of the claim.
The idea is to autogenerate all of the backend and the service layer for this including the UI. The grunt work of defining the screens etc., would be automated. The programmer would focus on building key business logic such as "claim processing upon submission".
Does this idea have merit?
The larger issue is that no system is an island, from a data perspective, and most work is brown, not green field. Subscribers and policies already live in the system the company has been using for 20 years- well, likely in a dozen systems that the business has acquired from competitors and continues to operate while struggling with data integration. Meanwhile marketing and sales has been using multiple generations of SAAS tools, finance has their own on prem or SAAS tools- all of which need to work with a common set of entities but which have different opinions about exactly how those entities are shaped. This is the reason data lakes and SAAS ETLs are so important right now.
Static/dynamic generators don't solve a painpoint at this point. It's not about new systems/data, it's about existing systems and data. HTH.