For the modern enterprise, experimenting with AI is over. Now, the focus has shifted from "what if" to "how fast," and as digital value chains evolve, CXOs are seeking to fundamentally reshape the economics of software development rather than just looking for increased productivity.
A report by Constellation Research on Thoughtworks’ AI/works™ platform highlights how the fabled goal of developer velocity has become a reality by injecting automation into every link of the digital value chain. With the AI/works™ Agentic Development Platform, teams can instantly decode legacy code, bridge the gap between business requirements and execution, and accelerate the entire build cycle.
But truly changing the economics of software delivery takes more than just adding AI plug-ins to old workflows. It requires rebuilding an AI-ready core that ensures security and scalability.
Moving beyond “bolted on” AI
Many organizations started to incorporate AI by adding features to traditional integrated development environments (IDEs). However, as the Constellation Research report notes, these "bolted-on" capabilities often lack a deep understanding of enterprise demands.
To change the cost and speed of development, enterprises are now moving toward AI-native platforms that don't just suggest the next line of code, but provide a framework for the entire software development lifecycle (SDLC). In this new model, the human role shifts from writing code to editing and validating it.
The power of spec-centric development
One of the most significant aspects of modern AI platforms is the return to specifications (specs). While code is often understood only by developers, specs can be read and updated by business leaders across the organization.|
By prioritizing spec-driven development, companies can bridge the gap between business intent and technical execution, across:
Code-to-spec: AI can now parse, deconstruct and synthesize legacy code bases, elevating them back to a "spec level” that humans can understand.
Spec enrichment: These specs can then be enriched with industry best practices and vertical-specific content.
Spec-to-code: Finally, the platform generates high-quality, spec-conforming code at previously unseen speeds.
This "spec-centricity" ensures that the software actually does what the business requires, which Constellation calls "the ultimate prize of software development".
Reshaping the economics
This economic shift in IT services is revolutionary. By leveraging AI-native platforms and a composable architecture of continuously updated context and components, an individual developer can multiply their capacity by 5x to 10x.
This unprecedented velocity reshapes enterprise strategy in two ways:
Flipping modernization economics: AI acceleration drastically compresses the transition window, breaking the financial gridlock of maintaining dual systems. Organizations move from static codebases to a stream of continuously updated, modernized code.
Evolving maintenance: As autonomous agents handle end-to-end tasks, the traditional maintenance cycle is replaced by software that constantly adapts to evolving user needs, UI choices and cyber defenses.
The bottom line is that optimizing the balance between human supervisors and AI platforms permanently drives down the cost of building and running digital assets.
Guidance for the C-suite
To rebuild for this new reality, executive leadership should consider three best practices:
Prioritize deterministic, orchestrated platforms: Look for foundational platforms/frameworks, like multi-agent orchestration and harness engineering, that complement and scale your existing tools — rather than replacing them with unproven tech.
Build, learn, then govern: While security is paramount, don't let rigid oversight stall progress. Instead, embrace adaptive governance — using guardrails that scale with usage rather than a blunt "no" up front.
Focus on value, not just cost: Evaluate technology by its business outcomes and compounding value, not just cost. By focusing on reliability, fast modernization, human-augmented AI and deterministic outputs, you can avoid technical debt and build a predictable, future-proof engine for growth.
Rebuilding the core isn't just about coding faster. It's about building a faster, smarter and more cost-effective engine for digital transformation.
The Constellation Research Pulse Report: Thoughtworks jolts enterprise AppDev into the AI era
Read the full report to discover how Al/works™ enables enterprises to achieve higher developer velocity through a spec-centric approach. Discover critical 2026 trends and get a deeper insight into Thoughtworks' unique ability to modernize legacy code into actionable specifications and high-quality automation.
Disclaimer: The statements and opinions expressed in this article are those of the author(s) and do not necessarily reflect the positions of Thoughtworks.