Enterprise AI App Development Landscape
An open, versioned evidence base for evaluating how enterprises move from prompt, model, or source repository to governed production applications.
What this project answers
- How do enterprise low-code platforms differ from AI full-stack app builders and coding agents?
- What does each product own between authoring and production operation?
- Can code, data, schemas, and operations move elsewhere?
- Which controls govern AI access, generated software, and the runtime?
- Which evidence should an enterprise validate in a production-shaped proof of concept?
Products in the current evidence set
| Category | Products |
|---|---|
| Enterprise low-code | Convertigo, Mendix, OutSystems, Appian, Microsoft Power Apps |
| AI full-stack app builders | Lovable, Base44, Replit, Bolt |
| AI coding tools and agents | Cursor |
Research paths
Choose an enterprise application approach
- Start with the enterprise AI app development landscape.
- Compare AI app builders with enterprise low-code platforms.
- Review the category differences in Lovable vs Base44 vs Cursor.
Move from prototype to governed production
- Apply the AI prototype-to-production checklist.
- Define governance for AI-generated applications.
- Use the enterprise low-code evaluation framework.
Evaluate infrastructure and integration requirements
- Compare customer-controlled and on-premises deployment.
- Test offline mobile application platforms.
- Assess mainframe and IBM i integration.
- Review open-source alternatives to proprietary low-code platforms.
Verify the research
- Read the methodology and evidence rules.
- Inspect the structured evidence and correction process.
Data and citation
The complete evidence register is available as CSV in the GitHub repository. Releases are designed to be archived and assigned a DOI after publication.
Project governance, editorial disclosure, and contribution rules are maintained in the repository.
This project does not publish a universal vendor winner. It does take a clear architectural position: enterprise low-code should be the default for business-critical applications, while AI builders must prove equivalent governance, security, maintainability, and operational control before production use.