vFunction brings GenAI-powered architecture tools to code assistants
vFunction has introduced new GenAI-powered features designed to bring architectural intelligence to code assistants such as Amazon Q Developer and GitHub Copilot, allowing for automated application refactoring and cloud-native transformation.
Prior to this release, code assistants operated without architectural context, limiting their capabilities to minor syntax improvements and preventing them from addressing broader modernisation efforts. The new features from vFunction intend to address this gap by making architectural context accessible to code assistants, facilitating system-wide improvements.
Architecture-aware functionality
The update includes architecture-aware prompts enabling code assistants to handle complex refactoring tasks. These tasks range from eliminating circular dependencies to decomposing large, monolithic classes, often referred to as "god classes". The prompts are generated in real time, using architectural context provided by vFunction's observability engine. This context draws on static and dynamic analysis as well as proprietary data science methodologies trained on actual customer applications.
A new MCP server has also been released, integrating vFunction's observability engine with developer command lines and integrated development environments (IDEs). Native integrations with Amazon Q Developer and GitHub Copilot allow developers to utilise these assistants more effectively for architectural remediation tasks.
GenAI-powered remediation
According to vFunction, the GenAI-powered features can automate remediation for various critical architectural issues, including the removal of dead code, domain dependency refactoring, and the extraction of services into discrete components. These tasks are streamlined using data sets derived from real-world applications, ensuring relevance and consistency.
"With these new advancements, teams can surface and resolve architectural debt, and transform their apps to cloud-native, with unprecedented speed through autonomous modernisation," said Amir Rapson, CTO and co-founder of vFunction.
"From eliminating circular dependencies to refactoring 'god classes', developers can now simplify refactoring and modernisation, accelerate delivery, and optimise architecture for the cloud."
Developer adoption of AI tools
Reference to the 2024 Stack Overflow Developer Survey is included in vFunction's announcement, highlighting that 76% of developers are using or plan to use AI tools in their development process. However, it notes that current GenAI tools have primarily focused on smaller tasks such as generating code snippets or automating repetitive actions, rather than large-scale refactoring with architectural awareness.
vFunction has developed architectural context-enriched prompts to ensure that GenAI assistants produce consistent and predictable outcomes, addressing the common problem of varying outputs from different AI-powered tools. These capabilities are now available to developers from within their workflows, whether in an IDE or at the command line.
MCP server integration
The new MCP server connects vFunction's observability insights directly to developer tooling, letting engineers query architectural issues, generate GenAI prompts, and initiate remediation steps without leaving their development environment. Integration with code assistants such as Amazon Q Developer and GitHub Copilot is available, with prompts informed by real-time data on the application's architecture.
vFunction lists several types of architectural issues that can now be addressed using GenAI prompts, including the removal of unnecessary code and dependencies, refactoring domain dependencies and utility class design patterns, splitting and re-architecting overgrown classes, and resolving circular flows that can impact application resilience and scalability.
"We're turning architectural observability into action," said Rapson. "Instead of just showing teams what's wrong with their applications, we're giving them AI-powered tools to fix these problems automatically. This bridges the gap between identifying architectural issues and resolving them to accelerate cloud modernization by orders of magnitude, taking the modernization process from months to days."
With the adoption of such tools, vFunction aims to support organisations planning to migrate to cloud-native architectures, helping to move beyond simple lift-and-shift cloud migrations by boosting the modularity and functionality of legacy applications. These tools are designed to facilitate faster and more consistent applicationmodernisationn within the context of existing development workflows.
