From Static Mocks to Design-in-Production
AI design tools are undergoing a fundamental shift: instead of living in isolated canvases, they are moving directly into production codebases. Platforms built around design-in-production let product teams prototype and explore using the same components, design systems and environments that power live applications. This approach tackles a long-standing disconnect. Designers and product managers have traditionally worked in mocks, documents and screenshots, while engineers operate in code and users only see the finished product in production. The translation between these worlds costs time, fidelity and ideas. By embedding design workflows inside real applications, teams can run product prototyping against actual logic, data and UI states. The result is tighter collaboration, fewer handoff errors and faster iteration, because every proposed change is grounded in how the software truly behaves, not how a static mock suggests it might.
Dessn’s Bet on Designing Inside Real Codebases
Dessn exemplifies this new wave of AI-assisted engineering. The platform lets designers and product managers prototype directly on top of their organisation’s real app, components and design system, without opening an IDE or running code locally. Instead of recreating interfaces in a separate tool, Dessn starts from the codebase and builds a design environment around it. This flips the traditional model and addresses what the company calls the “localhost problem” – the friction and setup required for non-engineers to access realistic product states. Dessn operates with read-only access, never writing or pushing code back, and each project runs in its own isolated microVM, an important assurance for teams wary of security and compliance risks. With product teams already dedicating several hours a day to the platform, Dessn is using its recent €5 million raise to expand a community of builders who want to explore every possible version of their product directly inside production.
DesignVerse Targets Legacy System Modernization
While some AI design tools focus on greenfield products, DesignVerse is built for legacy system modernization in complex enterprise environments. Its platform generates software using an organisation’s existing design systems, component libraries, technical documentation and internal rules. By doing so, it aligns new applications with current architectures and engineering standards rather than forcing teams to retrofit generic AI-generated code. This matters in sectors like aviation, finance, cybersecurity and government, where reliability, compliance and security are non-negotiable. DesignVerse aims to reduce the manual translation between design and engineering teams that often slows large projects, allowing stakeholders to validate behaviour earlier and streamline the transition from design to production. Backed by more than $5.5 million (approx. RM26 million) in seed funding, the company plans to expand its engineering team and accelerate adoption across enterprise markets that need AI-assisted engineering without compromising existing infrastructure.
Enterprise Modernisation Meets AI-Assisted Engineering
The rise of design-in-production platforms reflects broader enterprise pressures to modernise software without destabilising mission-critical systems. General-purpose AI coding tools can spin up quick prototypes, but integrating those outputs into production environments remains risky. By contrast, platforms like Dessn and DesignVerse are purpose-built to sit on top of real codebases and design systems, ensuring that product prototyping and implementation share the same source of truth. For modernisation projects, this means updates can be explored safely within the constraints of existing architectures and governance. Teams can experiment with new flows, validate them with stakeholders and then move to implementation with fewer surprises for engineering. As enterprises look to refresh legacy stacks and accelerate delivery, design-in-production approaches are emerging as a pragmatic middle ground: they preserve the rigour of established workflows while giving product teams the creative latitude to explore the “space of possibilities” that AI-driven tools unlock.