Assisted Low-Code Generation and Optimization: Bridging AI and App Design

Organisations involved

Main Participant: WebRatio is an Italian SME with more than 20 years of innovation in enterprise software development and the creator of the Interaction Flow Modelling Language Object Management Group® standards development organization. Its low‑code platform now integrates AI‑assisted development for web, mobile and process‑driven applications.

Technology Expert: RES IT is an established specialist in AI, Machine Learning and Deep Learning, providing proprietary software and data‑driven solutions to enterprise and financial customers.

The challenge

Enterprises face increasing pressure to accelerate digital transformation despite persistent shortages of skilled developers and growing application backlogs. Traditional coding is too slow and costly, while existing low-code platforms still require specialist expertise. Generative AI can help, but without domain-specific understanding, it often lacks the reliability and control needed for enterprise-grade solutions.

This creates two urgent needs: professional developers must boost productivity without compromising quality, and business users must gain greater autonomy to build applications with fewer IT bottlenecks. At the same time, major low-code vendors from the United States are rapidly adding AI-driven features, placing competitive pressure on European SMEs to innovate at speed.

WebRatio needed to act immediately to maintain feature parity, protect market share and open new revenue opportunities. Its strategy was to integrate a proprietary AI Low-Code Copilot capable of turning natural-language inputs into high-quality application models, enabling entry into the fast-growing Citizen Developer segment and doubling its addressable market.

To achieve this, WebRatio trained a Large Language Model to understand its domain-specific language (IFML) using more than 20 years of archive data. The required computational power made this financially unfeasible without HPC resources. This enabled large-scale training, ensuring model convergence, technical feasibility and timely innovation.

The Solution

WebRatio developed Assisted Low-Code Generation and Optimisation (ALGO) - an AI-powered copilot integrated into its Low-Code Development Platform. ALGO accelerates development by up to 50% for professional developers by automating repetitive modelling tasks, while lowering technical barriers for non-experts through natural-language interaction and guided model generation.

More than 100,000 archive samples were incorporated into a high-quality training dataset, which was used to fine-tune and benchmark multiple open-weight LLMs. The best-performing model was then embedded into a graph-based framework equipped with syntactic and semantic guardrails. This architecture ensures that each natural-language prompt is interpreted accurately, refined through domain rules, and converted into a fully executable, high-quality application model.

Impact 

ALGO demonstrates that generative AI can safely automate model creation within enterprise‑grade low‑code tools. For business, it delivers a dual impact: significantly higher productivity for professional developers and a new addressable market of citizen developers, enabling broader adoption within organisations. 

Socially, the solution democratizes software creation by enabling non‑technical users to build functional applications, enhancing digital inclusion and strengthening collaboration between business and IT teams. 

Strategically, the project shortens time‑to‑market for digital services and reinforces European competitiveness by delivering a trustworthy, EU‑developed alternative in a market dominated by non‑EU vendors. It also provides the foundation for WebRatio’s long‑term AI adoption roadmap, reinforcing the company’s role as an innovation-driven SME and contributing to a stronger European ecosystem for trustworthy, domain-specific AI.

Benefits

• 50% productivity gain for professional developers through AI-assisted modelling.
• ~€40,000 annual value created per developer for customers.
• 100% expansion of the addressable market by enabling citizen developers.
• Projected €2.1M in new annual recurring revenue by 2030 from AI-enhanced subscription tiers and licenses.