‘Software will eat software’ changes the economics of cutting your technical debt

Internet pioneer and VC Mark Andreessen once opined that software would eat the world, transforming businesses’ operations and creating new growth opportunities. Now, Mphasis CEO Nitin Rakesh believes that enterprise leaders can modernize legacy systems more affordably while updating processes, thanks to a new approach in which ‘software will eat software.’

His concept aligns well with HFS Research’s recently developed Services Tech Vision 2030 (Exhibit 1), in which we see AI-led agentic services emerging—augmenting human capabilities with intelligent AI agents to enhance service delivery. We expect agentic services to be followed by Service-as-a-Software—a technology-led service model that leverages advanced software solutions to provide technology arbitrage.

Exhibit 1: Software is becoming the new services

Source: HFS Research, 2024

Mphasis Neo platform services take strides toward ‘software will eat software’

As Nitin explained to analysts and advisors at the firm’s event at Lord’s Cricket Ground in London, Mphasis is already making strides toward the concept of ‘software will eat software’—best demonstrated by its NeoZeta and NeoCrux platforms.

Rather than simply copy and paste a rewrite of legacy code with generative AI (GenAI), these platforms use GenAI to understand what the legacy software does and how it does it (including where the data is accessed and stored for it to function). This supports the generation of a knowledge base from which domain-trained LLMs using Mphasis RAG engineering generate epics enriched with user stories, business logic ingrained in the code, database schema, and user interface elements.

These user stories become the interface with the business. They can help SMEs make decisions about stories that are no longer required and identify gaps for new requirements. The platform also supports the reduction of technical debt in modernized applications by reporting defunct, deprecated, and unused code in the legacy platform. Transforming legacy code into these specifications allows the modernization process to be independent of the choice of target technologies.

Modernize and add new requirements at the same time

Thanks to the libraries of domain and function-specific code Mphasis has developed, legacy platforms can be modernized and updated to align with new requirements—all in the same process.

The software conducts many steps—which previously would have required the time, effort, and extensive services of business analysts, legacy platform specialists, and engineers—using a combination of AI LLM models trained in different domains, a prompt engineering layer, and humans in the loop. Mphasis claims 50% savings in ‘relearning’ time by unleashing enterprise knowledge hidden in code and documentation and up to 60% cost and time savings in the price per line of code modernization.

Mphasis NeoCrux is an AI-powered platform for software engineers, streamlining the software development lifecycle (SDLC) using GenAI. It offers 40 capabilities for repurposing enterprise investments, from design to implementation and proactive monitoring. Mphasis claims a 40% increase in software productivity through code generation and test automation. This leads to the launch of new products and services up to 60% faster, along with ongoing insights and feedback for improvements.

Software-led services change the economics

The level of savings referenced above encourages fresh modernization efforts even where the business case was previously hard to make. Nitin provided an example of a transactions provider client responsible for around 50% of all US merchant transactions. While the firm’s code was half a century old, it was reliable and fast. However, the legacy code constrained the firm to one code release a year—and it could take 18 months to onboard a new merchant.

The new economics of emerging tech arbitrage-led solutions make the innovation benefits of modernizing your code viable where the numbers did not stack up in previous labor-intensive service solutions.

The Bottom Line: Reconsider the cost-benefit analysis of cutting your technical debt—the balance has shifted.

Technology arbitrage is cutting the cost of modernization. Software-led delivery platforms such as Mphasis’ NeoZeta and NeoCrux are revolutionizing how enterprises get work done. Every enterprise leader burdened by technical debt should revisit their assumptions about the economics of leaving legacy behind for a brighter and more innovative future.