Why nAG’s Interface System Matters Today: Inside the Engineering


Published 24/03/2026 By nAG

This blog is third in the series Making Numerical Algorithms Truly MultiLanguage. Read blog one and two.

The XML/XSLT system developed for the nAG Library solved a complex development problem: maintaining thousands of numerical algorithms across multiple languages. 

But as organisations increasingly want to expose algorithms — internally or to partners — the value of this system is broader than ever. 

Why the World Needs MultiLanguage Algorithm Delivery 

Across industries, we see the same trend: 

  • AI and ML teams exposing optimization or simulation kernels 
  • Engineering organisations distributing internal modelling algorithms 
  • Research groups publishing toolkits for broad adoption 
  • Enterprises needing APIs for Python, C#, Java, and the cloud at the same time 

These teams all face the same question: 

How do we make our algorithms available in many environments without multiplying work or creating inconsistencies? 

Most reinvent the wheel — handwritten wrappers, fragile documentation, diverging interfaces. 

nAG solved this problem years ago, at industrial scale. 

What nAG’s Approach Delivers

The XML/XSLT system provides: 

✔ A single source of truth 

No drift. No inconsistency. 

✔ Automatically generated wrappers 

Faster release cycles, fewer bugs, easier onboarding. 

Consistent Documentation  

Documentation always matches the code because they come from the same file. 

Portability Across Languages  

Once the pipeline is in place, adding a new language is a small, contained piece of engineering work. 

A Proven Track Record  

This system has supported one of the world’s most widely used numerical libraries and continues to do so today. 

A Strategic Opportunity 

For organisations looking to: 

  • distribute domain-specific algorithms 
  • expose internal scientific kernels 
  • modernise ageing toolkits 
  • create cloud-ready or AI-ready numerical APIs

nAG’s long-standing technical expertise in structured interfaces and automatic code generation offers a powerful foundation. 

This is a capability that very few organisations have — and one that is suddenly more relevant than ever.